Configure metadata (Metadata Management app)¶

Note

This section documents the Metadata Management app, which replaces the functionality of the Maintenance app. The Metadata Management app is available from the DHIS2 App Hub and is compatible with DHIS2 core V41 and above.

If you are using DHIS2 without the Metadata Management app installed, see Configure metadata for the Maintenance app instead.

About the Metadata Management app¶

In the Metadata Management app you configure all the metadata objects you need to collect and analyze data:

• Table: Dimensions and dimension items
• startDate 和 endDate 参数允许获取链接的数据 到这些日期之间的任何时间段。这避免了定义所有 期间明确在 要求:
• 数据集和数据输入表格
• Indicators and Predictors
• E7229
• CATEGORY_OPTION_GROUP
• Validations
• createdAtClient
• 常数
• 传说
• SQL views
• Locales
• Analytics table hooks

Note

The functions you have access to depend on your user role's access permissions.

The app's sidebar also has Programs and Data approval sections, which this chapter does not cover. For those, see Configure programs in the Maintenance app and Data approval overview.

Note

The Metadata Management app is released on a continuous release cycle, meaning new sections and features are added regularly. If a section listed above is not yet visible in your installed version of the app, it may not yet have been implemented. You can update the app from the App Hub to get the latest features.

The Metadata Management app is a complete modernisation of the legacy Maintenance app. Key improvements include:

• Merging duplicate metadata — You can merge duplicate metadata objects (currently: category options, indicators, indicator types) directly from the list view. This supports ongoing metadata cleanup.
• Advanced filtering and search — Each object list has improved filtering options to help you find objects by name, code, or other attributes.
• Bulk sharing — You can update sharing settings for multiple objects at once.

Navigating metadata objects¶

The left-hand sidebar lists all available metadata sections. Click any section to expand it and see the individual object types it contains, as shown below.

When you open the Metadata Management app, the left-hand sidebar displays the metadata object categories. Click a category to expand its object types, then click an object type to view its list.

Metadata objects are presented in a list with predefined columns relevant to each object. You can customise which columns are shown in the list for the current object. These customisations are per user and will not affect other users. Note that these changes do not edit any metadata — they only affect how the list is presented.

管理可见列¶

1. Click the Manage View button above the list of objects you want to configure.
2. A dialog appears with two tabs: Columns and Filters. Select the Columns tab.
3. The default selected columns are shown on the right.
4. Click any column name in the list of Available columns on the left to move it to the selected list.
5. Use the arrow buttons to move columns between the Available and Selected lists.
6. Reorder selected columns using the up and down arrow buttons.
7. Click Update view once you are satisfied with your changes.

Click Reset to default to restore the default column configuration.

Filtering and searching¶

Each object list includes a search bar and a row of filter dropdowns at the top of the list. The available filters are specific to each object type — for example, the data elements list can be filtered by Domain type, Value type, Data set, Category combo and Data element group.

You can also customise which filters appear using the Manage View button:

1. Click Manage View above the list.
2. Select the Filters tab.
3. Move filters between Available filters and Selected filters as needed.
4. Click Update view.

Download metadata¶

You can download the metadata for the object you are currently viewing. The download will respect any active filters.

1. Click the Manage View button above the list.
2. In the dropdown, select Download.
3. In the dialog that appears, select your preferred format and compression.
4. Optionally select With sharing to include sharing data for the metadata.
5. 点击**下载**。

All Metadata view¶

The new All metadata view allows you to search for keywords across a range of metadata types. Typing in the search box filters the results automatically, and the metadata types dropdown lets you narrow down which types of metadata are displayed. Expanding a type shows five results initially, with the option to load more.

Bulk update sharing¶

In the Metadata Management app you can select multiple entries in the list view and update the sharing settings of all of them at once. First select the items you want to update. In the example below, the view has been filtered to show only category options in the category Labour complications, and all of them have been selected.

Clicking Update sharing opens a dialog where you choose which sharing settings to add to the selected objects.

Further options for editing objects' existing sharing settings are under development.

Common metadata object fields¶

Many metadata objects in the sections below share the same handful of fields. This table explains what each one is for and any format constraints, so the individual procedures don't have to repeat it. Where a step below names one of these fields with no further explanation, this table is the reference — sections only add their own explanation for fields that are unique to that object type.

Table: Common metadata object fields

| The manual merge is suitable when there are resolvable conflicts or when not all the data needs to be moved during the merge. For example, if an attribute has different values in both tracked entities , the user can specify whether to keep the original value or move over the duplicate's value. Since the manual merge involves the user explicitly requesting to move data, there are some additional checks: | Used on | Purpose and best practice | | --- | --- | --- | | 类型 | All objects | The full display name shown throughout the app, in lists and in most reports. Must usually be unique within the object type. | | 稳定的标识符。最多 11 个字符。如果未指定,将由系统生成。 | Most objects | A shorter version of Name, often with a character limit (commonly 50, sometimes fewer — the form tells you if you exceed it). Used wherever space is tight, for example Pivot Table column headers or mobile apps. | | zscore | All objects | An optional identifier made of letters, numbers, dots, dashes and underscores (no spaces). Codes are how external systems and data exchanges reference DHIS2 objects, so changing a code once it's in use elsewhere can break integrations. Agree on a coding convention before creating objects at scale rather than retrofitting one later. | | 简称 | Data elements, category options and similar objects | An alternative name shown as a column header in data entry forms instead of the display Name. Use it when the display Name is too long or phrased differently from what should appear on a data entry form. | | Parameter name | All objects | Worth filling in even where marked optional. Explain what the object measures or represents and why, in enough detail that someone unfamiliar with the project could understand it without asking you. | | Start date / End date | Category options and other time-bound objects | Restricts when the object is available, for example a category option that only applies during a specific project phase. | | Color / Icon | Data elements, indicators, options | Optional visual styling shown in data capture apps (Capture, DHIS2 Android Capture) to help users tell objects apart at a glance. | | Data dimension / Data dimension type | Categories, category combinations, category option groups and sets, data element and indicator group sets | Controls whether the object becomes available as a selectable dimension in analytics apps such as Pivot Table and Data Visualizer, alongside built-ins like Period and Organisation unit. Where the object offers a choice such as Disaggregation vs Attribute, that choice is specific to how the object is used — see the object's own section below. |

Common actions¶

Show details, Sharing settings, Translate, Delete and Clone work the same way for almost every object type covered in this chapter, regardless of whether it's a category, a data element, a predictor or something else. This section documents each action once. Each object type's own section below only notes where it deviates from this — for example an object type that doesn't support one of these actions, or has an object-specific detail worth knowing before you use it.

Show details¶

1. Open the Metadata Management app and find the type of metadata object you want to view.
2. 在对象列表中，单击选项菜单并选择**显示详细信息**。

Change sharing settings¶

You can assign different sharing settings to metadata objects to control which users and user groups can view or edit them, for example organisation units and tracked entity attributes.

Some metadata objects also allow you to change the sharing setting of data entry for the object. These additional settings control who can view or enter data in form fields using the metadata.

Note

The default setting is that everyone (Public access) can find, view and edit metadata objects.

1. Open the Metadata Management app and find the type of metadata object you want to modify.
2. In the object list, click the options menu and select Sharing settings.
3. (Optional) Add users or user groups: search for a user or user group and select it.
4. Change sharing settings for the access groups you want to modify:
5. Can edit and view: The access group can view and edit the object.
6. 只能查看：访问组可以查看对象。
7. No access (only applicable to Public access): The public will not have access to the object.
8. Change data sharing settings where applicable:
9. Can capture data: The access group can view and capture data for the object.
10. Can view data: The access group can view data for the object.
11. No access: The access group will not have access to data for the object.
12. 点击**关闭**。

Translate¶

DHIS2 provides functionality for translations of database content. You can translate a metadata object into any number of locales. A locale represents a specific geographical, political, or cultural region.

Tip

To activate a translation for yourself, open the System Settings app, click Appearance and select a language.

1. Open the Metadata Management app and find the type of metadata object you want to translate.
2. 在对象列表中，单击选项菜单并选择**翻译**。

Tip

Organisation unit levels don't have an options menu — click directly on the Translate icon next to the list item instead.

1. 选择一个语言环境。
2. Type a Name and optionally a Short name and Description in the chosen locale.
3. 点击**保存**。

Delete¶

Note

You can only delete a metadata object if no data is associated with it.

Where deletion is blocked, or where you would rather retire an object than lose its history, use Sharing settings to restrict access instead. The object and its data still exist, but users without access won't see them. Some object types offer a dedicated alternative — an organisation unit can be closed rather than deleted, for example.

Warning

Deleting objects from DHIS2 is not recommended. In almost all cases it is better to deactivate or close an object rather than deleting it, as deletion is irrevocable.

1. Open the Metadata Management app and find the type of metadata object you want to delete.
2. 在对象列表中，单击选项菜单并选择**删除**。
3. 点击**确认**。

One object type carries an extra risk beyond the general note above:

• Data sets — deleting a data set also irrevocably deletes any data entry forms and section forms built for it. Make sure you have a backup of your database before deleting a data set, in case you need to restore it later.

Clone¶

Cloning a metadata object can save time when you need to create many similar objects.

1. Open the Metadata Management app and find the type of metadata object you want to clone.
2. 在对象列表中，单击选项菜单并选择**克隆**。
3. All values from the original object are copied. Values that must be unique, such as Name and Code, need to be updated. Modify the other values you want to change.
4. 点击**保存**。

Clone is available for most object types in this chapter. It is not available for: Category option combination, Data set, Section, Section form, Custom form, and Organisation unit level.

Merge¶

Merging consolidates two duplicate objects into one: it reassigns all references from a source object to a target object, then deletes the source. Unlike the actions above, merge is available for only some object types, and the navigation path and required authority differ by type.

The Metadata Management app provides a merge screen for eight object types:

All eight work the same way: open the object type's list, choose Merge, select the source object or objects to be removed and the target object to keep, review the summary of changes, then confirm. Each merge screen requires its own authority, and merging is irreversible — the sources are deleted and every reference is repointed at the target. Three are written out in full below; the rest follow the same pattern from their own list view.

Using a transfer list component¶

Many sections below ask you to assign one type of object to another — category options to a category, say, or data elements to a data element group — using the two-panel selector shown below. Wherever a procedure tells you to "select X and assign them", this is the component it means.

Available (left) panel

• Lists every object of that type that could be assigned, whether or not it's already assigned elsewhere.
• Some object types add a Show only unassigned items checkbox above this list, which narrows the choice to items that do not already belong to another object of the same type. It is most useful when assigning members to groups.
• Each panel also has its own filter box above the list, so you can search the Available and Selected lists independently.

Moving items to Selected

• Double-click a single item in the Available list to move it to the Selected list immediately.
• Click an item to select it, or hold Shift or Ctrl while clicking to select a range or several individual items, then click the single right-pointing arrow to move the selection across.
• Click the double right-pointing arrow to move every item in the Available list to Selected, regardless of what's selected.

Removing items from Selected

• Double-click a single item in the Selected list to move it back to Available.
• Click an item to select it, or hold Shift or Ctrl while clicking to select a range or several individual items, then click the single left-pointing arrow to move the selection back.
• Click the double left-pointing arrow to move every item in the Selected list back to Available.

Add new

Where a transfer list offers an Add new button, clicking it opens a drawer with a blank form for creating a new item. Complete the form and click Save and close. The new item then appears in the Available list, ready to be assigned. This lets you create a missing item without leaving the parent object you are editing.

Manage categories¶

About categories¶

类别通常是一个概念，例如“性别”，“年龄”或“ “疾病状态”。数据元素，例如“已确认病例数 疟疾”通常分为较小的组成部分，以便确定 例如，已确认的特定年龄组的疟疾病例数。

Use categories to disaggregate data elements into individual components. You can also use categories to assign metadata attributes to all data recorded in a specific data set, such as "Implementing partner" or "Funding agency."

For example, create a category called "Age" with three category options: "Under 1", "1-5" and "Over 5". Assign that category to the data element through a category combination. This creates three separate fields for the data element in the data entry forms:

• 确诊的疟疾病例数（1岁以下）
• 确诊的疟疾病例数（1-5）
• 确诊的疟疾病例数（5例以上）

没有类别，您将不得不创建每个数据 上面单独列出的元素。

In the Metadata Management app, you manage the following category objects:

Workflow¶

1. 创建所有类别选项。
2. Create categories composed of the multiple category options you have created.
3. Create category combinations composed of either one or multiple categories.
4. 创建数据元素并将其分配给类别组合。

Create or edit a category option¶

When possible, reuse category options. For instance, there might be two categories which share a particular category option (for example "< 1 year of age"). When creating the categories, this category option can be reused. This is important if particular category options (or category option combinations) need to be analysed together.

1. Open the Metadata Management app and click Categories > Category options.
2. Click the + New button to add a new category option, or click the options menu next to an existing one and select Edit.
3. Fill in Name, Short name, Code, Form name, Description, Start date and End date — see Common metadata object fields. Only Name is required.
4. 选择组织单位并分配它们。

Tip

You can automatically select all organisation units that belong to an organisation unit level or organisation unit group, for example "Chiefdom" or "Urban". Select an Organisation unit level or Organisation unit group and click Select.

1. 点击**保存**。

Create or edit a category¶

为特定类别创建所有类别选项后， 您可以创建该类别。

1. Open the Metadata Management app and click Categories > Categories.
2. Click the + New button.
3. Fill in Name, Short name, Code and Description — see Common metadata object fields.
4. Select a Data dimension type — Disaggregation for a category used to break data elements into components, or Attribute to let the category be used to assign a combination of categories to data recorded through a data set.
5. (Optional) Select Data dimension to make the category available as a dimension in analytics apps.
6. 选择类别选项并分配它们。
7. 点击**保存**。

Create or edit a category combination¶

Category combinations let you combine multiple categories into a related set.

您可以分解数据元素“新感染艾滋病毒的数量” 分为以下几类：

• 艾滋病毒服务：“其他”，“ PMTCT”，“结核病”
• 性别：“男”，“女”

In this example, there are two levels of disaggregation that consist of two separate data element categories. Each data element category consists of several data element category options.

在DHIS2中，根据 常见的一组类别。通过将这些不同的类别合并为一个 类别组合并将这些组合分配给数据元素， 您可以将适当的分类级别快速应用于大型 数据元素数。

1. Open the Metadata Management app and click Categories > Category combinations.
2. Click the + New button.
3. Fill in Name and Code — see Common metadata object fields.
4. Select a Data dimension type — Disaggregation or Attribute.
5. (Optional) Select Skip category total in reports.
6. 选择类别并分配它们。
7. 点击**保存**。

Create or edit a category option group¶

您可以使用类别选项对类别选项进行分组和分类 组。类别选项组集的主要目的是添加更多 您捕获的数据的维数，以便在例如 **数据透视表**或**数据可视化器**应用。

Consider a system where data is collected by "projects", and projects are modelled as category options. The system must be able to analyse data based on which donor supports the project. In this case, create a category option group set called "Donor". Each donor can be created as a category option group, where each category option / project is put in the appropriate group. In the data analysis applications, the "Donor" group set appears as a data dimension and each donor appears as a dimension item, ready to be included in reports.

Category option groups allow you to classify category options into groups that work together.

1. Open the Metadata Management app and click Categories > Category option groups.
2. Click the + New button.
3. Fill in Name, Short name, Code and Description — see Common metadata object fields.
4. Select a Category option group usage — Disaggregation or Attribute. This works the same way as Data dimension type; see Common metadata object fields.
5. 选择类别选项并分配它们。
6. 点击**保存**。

Create or edit a category option group set¶

您可以将类别选项组分组在类别选项组集中。的 类别选项组集的主要目的是添加更多 您捕获的数据的维数，以便在例如 **数据透视表**或**数据可视化器**应用。

1. Open the Metadata Management app and click Categories > Category option group sets.
2. Click the + New button.
3. Fill in the Name and Description fields — see Common metadata object fields.
4. Select a Data dimension type.
5. Select category option groups and assign them.
6. 点击**保存**。

Use category combinations for data sets¶

If you use a category combination for a data set, the category combination applies to the whole form. This requires the categories and the category combination to have the data dimension type "Attribute". Used as an attribute, a category combination serves as another dimension — similar to "Period" and "Organisation unit" — which you can use in your analysis. When a data set is linked to a category combination, those categories are displayed as drop-down boxes in the Data entry app, and each data value captured in the form is linked to the category options selected in those boxes.

Suppose that an NGO is providing ART services in a given facility. They would need to report each month on the "ART monthly summary", which would contain a number of data elements. The NGO and project could potentially change over time. In order to attribute data to a given NGO and project at any point in time, you need to record this information with each data value at the time of data entry.

1. 使用数据维度类型“属性”创建两个类别： “实施伙伴”和“项目”。

2. 创建具有数据维类型的类别组合 “属性”：“实施合作伙伴和项目”。

3. 将您创建的类别分配给类别组合。

4. 创建一个称为“ ART每月摘要”的数据集，然后选择 “实施合作伙伴和项目”类别组合。

When you enter data in the Data entry app, you can select an "Implementing partner" and a "Project". Each recorded data value is assigned a specific combination of these categories as an attribute. These attributes (when specified as a dimension) can be used in the analysis applications similar to other dimensions, for example the period and organisation unit.

See the Manage data sets section for more on assigning a category combination to a data set.

Assign a code to a category option combination¶

The system creates the category option combinations automatically, so they do not need to be created. You can assign a code to category option combinations. This makes data exchange between DHIS2 and external systems easier.

1. Open the Metadata Management app and click Categories > Category option combinations.
2. In the object list, click the options menu and select Edit.
3. In the Code field, type a code.
4. 点击**保存**。

Merge category options¶

Merging allows you to consolidate two duplicate category options into one, reassigning all references from the source object to the target object and then deleting the source.

Note

You must have the F_CATEGORY_OPTION_MERGE authority to access the merge page or see the merge button on the list view.

1. Open the Metadata Management app and click Categories > Category options.
2. Navigate to Category options > Merge Category options (or go directly to #/categoryOptions/merge).
3. Search for and select the source category option (the one to be removed).
4. Search for and select the target category option (the one to keep).
5. Review the summary of changes.
6. Click Confirm merge.

Warning

Merging is irreversible. The source category option will be permanently deleted and all references will be updated to point to the target.

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions. All category objects support these except Category option combination, which only supports Edit and Show details — it's generated automatically by the system rather than created directly. Categories, category combinations, category option combinations and category options all support merging; see Merge.

Manage data elements¶

About data elements¶

数据元素是DHIS2的基础。数据元素定义什么 实际记录在系统中，例如免疫接种次数或 疟疾病例数。

诸如“确诊疟疾病例数”之类的数据元素通常 分解成较小的组成部分，例如确定 特定年龄段已确诊的疟疾病例数。

In the Metadata Management app, you manage the following data element objects:

Workflow¶

1. 创建所有类别选项。
2. Create categories composed of the multiple category options you have created.
3. Create category combinations composed of either one or multiple categories.
4. 创建数据元素并将其分配给类别组合。

Create or edit a data element¶

1. Open the Metadata Management app and click Data elements > Data elements.
2. Click the + New button to add a new data element, or click the options menu next to an existing one and select Edit.
3. In the Name field, define the precise name of the data element. Each data element must have a unique name.
4. In the Short name field, define a short name for the data element. Typically, the short name is an abbreviation of the full data element name. This attribute is often used in reports where space is limited.
5. （可选）在 Code 字段中，分配一个代码。
6. (Optional) In the Color field, assign a color which will be used for this data element in data capture apps.
7. (Optional) In the Icon field, assign an icon which will be used for this data element in data capture apps.
8. In the Description field, type a description of the data element. Be as precise as possible and include complete information about how the data element is measured and what its purpose is.
9. (Optional) In the Field mask field, type a template that provides hints for correct formatting of the data element.

Note

Field masks are currently implemented in the DHIS2 Android Capture app; not in the Capture web app.

The following special characters can be used in the mask:

For example, the pattern \d\d\d-\d\d\d-\d\d\d would show a hyphen for every third digit.

1. In the Form name field, type an alternative name of the data element. This name can be used in either section or automatic data entry forms.
2. In the Domain type field, select whether the data element is an aggregate or tracker type of data element.

3. In the Value type field, select the type of data that the data element will record.

   Value types

   Value type	Description
   Text	Textual value. The maximum number of allowed characters per value is 50,000.
   Long text	Textual value. Renders as a text area with no length constraint in forms.
   Letter	A single letter.
   Phone number	Phone number.
   Email	Email address.
   Yes/No	Boolean values rendered as drop-down lists in data entry.
   Yes only	Boolean values rendered as check boxes in data entry.
   Date	Dates rendered as calendar widget in data entry.
   Date & time	Combination of the Date and Time value types.
   Time	Time rendered as hours and minutes in HH:MM format.
   Number	Any real numeric value with a single decimal point.
   Unit interval	Any real number greater than or equal to 0 and less than or equal to 1.
   Percentage	Whole numbers inclusive between 0 and 100.
   Integer	Any whole number (positive and negative), including zero.
   Positive integer	Any whole number greater than (but not including) zero.
   Negative integer	Any whole number less than (but not including) zero.
   Positive or zero integer	Any whole number greater than or equal to zero.
   Username	DHIS2 user. Renders as a dialog for searching for and selecting a user.
   Coordinate	A point coordinate specified as longitude and latitude in decimal degrees. All coordinates should be specified in the format [-19.23, 56.42].
   Organisation unit	Organisation units rendered as a hierarchy tree widget. If the user has assigned "search organisation units", these will be displayed instead of the assigned organisation units.
   Reference	Stores a reference (UID) to another DHIS2 object, allowing a data value to link to a specific record elsewhere in the system.
   Age	Dates rendered as calendar widget OR by entering number of years, months and/or days which calculates the date value based on current date.
   URL	Enables manual entry of a URL.
   File	A file resource where you can store external files, for example documents and photos.
   Image	A file resource where you can store photos. Unlike File, the Image type can display the uploaded image directly in forms.
   GeoJSON	Stores geographic geometry data in GeoJSON format, for example boundary polygons or catchment areas for organisation units.

4. In the Aggregation type field, select the default aggregation operation that will be used for the data element.

   Table: Aggregation operators

   Every aggregation type answers two separate questions: how the shorter periods inside the reporting period combine, and how child organisation units combine into their parent. Read the names as "what happens over time (what happens over organisation units)" — where the organisation unit half is not stated, it is Sum.

   Aggregation operator	Over periods	Over the organisation unit hierarchy
   Sum	Sum	Sum
   Average	Average	Average
   Average (sum in org unit hierarchy)	Average	Sum
   Count	Count of values	Sum
   Min	Minimum	Minimum
   Max	Maximum	Maximum
   Min (sum in org unit hierarchy)	Minimum	Sum
   Max (sum in org unit hierarchy)	Maximum	Sum
   First value	Earliest value in the period	Sum
   First value (average in org unit hierarchy)	Earliest value in the period	Average
   Last value	Most recent value in the period	Sum
   Last value (average in org unit hierarchy)	Most recent value in the period	Average
   Last value in period	Most recent value, counting only periods inside the requested period	Sum
   Last value in period (average in org unit hierarchy)	Most recent value, counting only periods inside the requested period	Average
   Standard deviation	Standard deviation (population-based)	Standard deviation
   Variance	Variance (population-based)	Variance
   None	No aggregation	No aggregation

   Two further entries are not aggregation behaviours at all:

   • Default — defer to the aggregation type implied by the data element's value type.
   • Custom — the aggregation is supplied elsewhere rather than by one of the operators above.

   Tip

   The two-axis split is what makes Sum and Average (sum in org unit hierarchy) give different district totals from the same facility data, and it is why Last value at district level gives you the sum of each facility's latest value rather than any one facility's. If a total looks wrong, work out which axis you actually meant.

   Note

   Choose Last value for data that is a running total or a snapshot rather than a count of events — stock on hand, staff in post, population. Summing snapshots across periods double-counts them.

5. In the Store zero data values field, select if you want to save zero values, or if the system should ignore them.

   By default, DHIS2 does not store zero values. If you have a data element where a zero value is meaningful, for example if you want to track the number of vaccine doses administered and zero is a valid entry, select this option.

6. (Optional) In the Category combination field, define which category combination the data element belongs to. This is the disaggregation applied to the data element.

7. (Optional) Assign the data element to one or more Data element groups.

8. 点击**保存**。

Create or edit a data element group¶

Data element groups allow you to group related data elements for reporting and analysis purposes.

1. Open the Metadata Management app and click Data elements > Data element groups.
2. Click the + New button.
3. Fill in the Name and (optionally) Short name and Code fields — see Common metadata object fields.
4. 选择数据元素并分配它们。
5. 点击**保存**。

Create or edit a data element group set¶

Data element group sets allow you to classify multiple data element groups.

1. Open the Metadata Management app and click Data elements > Data element group sets.
2. Click the + New button.
3. Fill in the Name and Description fields — see Common metadata object fields.
4. If you want to use the data element group set in analytics, select Data dimension.
5. 选择数据元素组并分配它们。
6. 点击**保存**。

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions.

Manage data sets and data entry forms¶

About data sets and data entry forms¶

All data entry in DHIS2 is organised in data sets. A data set is a collection of data elements grouped together for data entry and data export between instances of DHIS2. To use a data set to collect data for a specific organisation unit, you must assign the organisation unit to the data set. Once you have assigned the data set to an organisation unit, that data set is available in the Data entry app. Only the organisation units you have assigned the data set to can use the data set for data entry.

A category combination can link to both data elements and data sets. If you use a category combination for a data set, the category combination is applicable for the whole form. This means that you can use categories to capture information which is common to an entire form, for example the name of a project or grant. When a data set is linked to a category combination, those categories are displayed as drop-down boxes in the Data entry app.

You create and edit data sets in the Metadata Management app. Here you define, for example, which data elements you want to include in the data set and the data collection frequency.

You enter data in the Data entry app. The Data entry app uses data entry forms to display the data sets. There are three types of data entry forms:

Note

If a data set has both a section form and a custom form, the system displays the custom form during data entry. Users who enter data cannot select which form they want to use. In web-based data entry the order of display preference is:

1. Custom form (if it exists)
2. Section form (if it exists)
3. Default form

Mobile devices do not support custom forms. In mobile-based data entry the order of display preference is:

1. Section form (if it exists)
2. Default form

In the Metadata Management app, you manage the following data set objects:

Workflow¶

You need to have data elements and categories before creating data sets and data entry forms.

1. 创建一个数据集。
2. Assign the data set to organisation units. A default form is created automatically.
3. 创建节形式或自定义形式。

Now you can register data in the Data entry app.

Create or edit a data set¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. Click the + New button.
3. In the Name field, type the precise name of the data set.
4. In the Short name field, type a short name for the data set. The short name must be unique.
5. （可选）在 Code 字段中，分配一个代码。
6. In the Description field, type a description of the data set.
7. In the Expiry days field, enter the number of days after which you cannot edit the data set. If set to 0, users can always edit.
8. In the Open future periods for data entry field, enter the number of future periods to make available for data entry.
9. In the Days after period to qualify for timely submission field, enter the number of days after the end of a period by which a report is considered submitted on time.
10. (Optional) In the Category combination field, select the category combination to use for the data set.
11. In the Period type field, select the frequency of data entry for this data set.
12. If applicable, in the Notification recipients field, select who should receive data set completion notifications.
13. (Optional) Select Complete allowed only if validation passes to prevent completing the data entry form if there are active validation rule violations.
14. (Optional) Select Complete allowed if compulsory fields are filled to prevent completing the data entry form until all data elements marked as compulsory (see Edit compulsory data elements in a data set) have a value.
15. (Optional) Select Skip offline if you want the data set to only be available when the user is connected to the Internet.
16. In the data elements section, search for and add the data elements to include in the data set.
17. Assign the data set to one or more organisation units.
18. 点击**保存**。

Create or edit a data set notification¶

Data set notifications are sent to notify users when a data set is completed or is about to expire. You can configure notification templates that determine what is sent, when, and to whom.

发送什么？¶

In the Metadata Management app you can configure a message template and a subject for the notification. The template can include variables such as {organisationUnit.name} and {period} to personalise messages.

何时发送？¶

You can configure notifications to send: * On completion — when the data set is marked as complete. * Scheduled days — a set number of days relative to the reporting period end date.

寄给谁？¶

Notifications can be sent to: * User groups * Organisation unit contacts

Override data elements' category combinations in a data set¶

You can override the category combination of a data element within a specific data set. This allows you to use a different disaggregation for the same data element in different data sets.

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, click the options menu and select Edit.
3. In the Data elements section, click the override option next to the data element whose category combination you want to override.
4. Select the replacement category combination.
5. 点击**保存**。

Edit compulsory data elements in a data set¶

You can mark specific data elements in a data set as compulsory. Users will be prompted before completing the data set if compulsory data elements have not been filled in.

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, click the options menu and select Edit compulsory data elements.
3. Select the data elements you want to mark as compulsory.
4. 点击**保存**。

Manage section forms¶

创建部分表格¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Click + New to add a new section.
4. Enter a Name for the section, and (optionally) a Code and a Description.
5. Select the data elements to include in this section.
6. Click Save section.

Note

Repeat for each section in the form. Each data element should belong to only one section.

编辑部分表格¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Click the section you want to edit.
4. Modify the section as needed.
5. Click Save section.

管理分区表单中的灰色字段¶

You can disable (grey out) specific cells in a section form to prevent data entry for particular combinations.

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Click Manage grey fields for the section you want to modify.
4. Click on the cells you want to grey out.
5. 点击**保存**。

在部分表格中更改部分显示顺序¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Drag and drop the sections to reorder them.

删除部分表格中的部分¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Click the delete icon next to the section you want to remove.

以节形式翻译节¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage sections.
3. Click the translate icon next to the section.
4. Select a locale and enter a translated name.
5. 点击**保存**。

Manage custom forms¶

Note

Custom forms use an HTML editor. While powerful, custom forms require more time to create and maintain than section forms.

创建一个自定义表格¶

1. Open the Metadata Management app and click Data sets > Data sets.
2. In the object list, open the relevant data set and navigate to Manage form.
3. Select Custom as the form type.
4. Use the HTML editor to design your custom form. Data elements can be inserted as input fields.
5. 点击**保存**。

自定义形式的脚本¶

Custom forms support JavaScript for advanced interactions and validation. Scripts are embedded directly in the HTML of the custom form.

大事记¶

职能¶

For Show details, Sharing settings, Translate and Delete, see Common actions — note the extra risk called out there for deleting a data set specifically. Data sets do not support Clone.

Manage indicators¶

About indicators¶

An indicator is a formula that can consist of multiple data elements, constants, organisation unit group counts and mathematical operators. Indicators typically consist of a numerator and denominator. You use indicators to calculate coverage rates, incidence and other values that are a result of data element values that have been entered into the system. Calculated totals do not have a denominator.

Note

You never enter indicator values directly in DHIS2 — they are always calculated.

An indicator formula can consist of mathematical operators (for example plus and minus), functions, and the following elements:

您可以在指标公式中使用以下功能：

Alongside these functions you can use:

• the mathematical operators +, -, *, /, % (modulus) and ^ (power), with parentheses to group parts of an expression
• the comparison operators ==, !=, >, <, >= and <=
• the logical operators && (or the keyword and), || (or or) and ! (or not)

Note

DHIS2 parses indicator, predictor, validation rule and program indicator expressions with one shared grammar, but each expression type only accepts its own subset of functions. A function that belongs to another type is recognised by the parser and then rejected when the expression is validated, which is why an expression can look plausible and still fail to save. In particular:

• Aggregation functions — avg(), sum(), count(), min(), max(), median(), stddev(), stddevPop(), stddevSamp() and percentileCont() — and the distribution functions normDistCum() and normDistDen() belong to predictor generator expressions. See Create or edit a predictor. To evaluate part of an indicator expression before it is aggregated, use subExpression instead, and to override how an item aggregates use .aggregationType().
• The orgUnit.ancestor(), orgUnit.dataSet(), orgUnit.group() and orgUnit.program() functions belong to validation rule and predictor expressions.
• The d2: functions and V{...} variables belong to program indicator expressions and program rules. See Reference information: Functions, variables and operators to use in program indicator expressions and filters.
• subExpression(), .aggregationType(), .periodOffset(), .yearToDate(), [periodInYear], [yearlyPeriodCount] and indicator references are accepted only in indicator expressions.

The aggregation types you can pass to .aggregationType() are:

Examples of .aggregationType, .maxDate, .minDate and .periodOffset:

Indicator SubExpressions¶

Normally an indicator aggregates a data element's values first and then evaluates the expression. For example, in the expression:

if( #{nYahlae7fe6} > 10, 1, 0 )

if the data element has aggregation type SUM, DHIS2 sums all of that data element's values for the relevant period and then tests whether the sum is greater than 10.

Sometimes you need the opposite: evaluate each value before it is aggregated. For example, to show how many facilities in a district recorded a value greater than 10, wrap the test in subExpression:

subExpression( if( #{nYahlae7fe6} > 10, 1, 0 ) )

Each individual data value is tested, returning 1 or 0, and (assuming the data element aggregates with SUM) the 1s and 0s are then summed — giving a count of how many values exceeded 10.

Notes on SubExpressions:

• The 1s and 0s are only summed if the data element's aggregation type is SUM. If it has a different aggregation type and you want them summed, override the aggregation type inside the sub-expression:

  subExpression( if( #{nYahlae7fe6} > 10, 1, 0 ) ).aggregationType(SUM)
  

• A sub-expression may reference only one data element, but it may reference it more than once. For example:

  subExpression( if( #{nYahlae7fe6} > 10 && #{nYahlae7fe6} <= 20, 1, 0 ) )
  

• A sub-expression may reference a data element with a category option combination and/or an attribute option combination, but it must be exactly the same reference each time. For example:

  subExpression( if( #{nYahlae7fe6.beec4Dewah8} > 10 && #{nYahlae7fe6.beec4Dewah8} <= 20, 1, 0 ) )
  

• To evaluate an expression before aggregation that involves program data, or more than one data element, category option combination or attribute option combination, use a predictor instead and store the result as a separate data element. You can then reference that predicted data element in an indicator or directly in analytics.

Indicator Year-to-date¶

Indicators can compute year-to-date values using three expression elements: .yearToDate(), [periodInYear] and [yearlyPeriodCount].

In the examples below, #{a} can be #{dataElementUID} or any valid indicator expression item that returns a data value, such as #{dataElementUID.catOptionComboUid}, I{programIndicatorUID} or N{indicatorUID}.

Note

A weekly period is treated as part of a year if four or more of its days fall in that year.

Notes on missing data¶

If data is missing for some of the periods in a year-to-date calculation, the result is based only on the periods that have data. Bear this in mind when interpreting year-to-date figures — a low value may mean under-performance or may simply mean a period has not been reported yet.

Workflow¶

1. Create indicator types (unless using the default types).
2. 创建指标。
3. (Optional) Create indicator groups.
4. (Optional) Create indicator group sets.

Create or edit an indicator type¶

Indicator types define the factor by which indicator values are multiplied when shown in reports or exported to analytics. A "Per cent" indicator type, for example, multiplies values by 100.

1. Open the Metadata Management app and click Indicators and Predictors > Indicator types.
2. Click the + New button.
3. In the Name field, type the name of the indicator type, for example "Per cent".
4. In the Factor field, type the factor. For a percentage indicator, type 100.
5. If the indicator type does not have a denominator (for example a count), select Number (without denominator).
6. 点击**保存**。

Create or edit an indicator¶

1. Open the Metadata Management app and click Indicators and Predictors > Indicators.
2. Click the + New button, or click the options menu next to an existing indicator and select Edit.
3. In the Name field, type the full name of the indicator, for example "Incidence of confirmed malaria cases per 1000 population".

4. In the Short name field, type an abbreviated name, for example "Inc conf. malaria per 1000 pop". The short name must be unique and 50 characters or less.

5. （可选）在 Code 字段中，分配一个代码。

6. (Optional) In the Color field, assign a color to represent the indicator.
7. (Optional) In the Icon field, assign an icon to illustrate the meaning of the indicator.
8. In the Description field, type a brief, informative description of the indicator and how it is calculated.
9. If you want to apply an annualization factor during calculation, select Annualized. This scales the numerator up to a yearly equivalent so it can be compared against a denominator that is already annual, such as a population figure. Typically the numerator is multiplied by 365 divided by the number of days in the reporting period.
10. Select the number of Decimals in data output.
11. Select an Indicator type. This field determines a factor that is automatically applied during calculation.
12. （可选）分配一个或多个**传奇**。
13. (Optional) In the URL field, enter a link — for example to an indicator registry — where a full metadata description of the indicator is available.
14. (Optional) Enter a Category option combination for aggregate data export and an Attribute option combination for aggregate data export. These settings are used when mapping aggregated data exported to another server.
15. (Optional) Enter values for any custom attributes.
16. Click Edit numerator:
    1. 键入分子的清晰描述。
    2. Define the numerator by double-clicking components in the right-hand field. Add mathematical operators by double-clicking the icons below the formula field.
    3. Click Done.
17. Click Edit denominator:
    1. 键入分母的清晰描述。
    2. Define the denominator using the formula editor as above.
    3. Click Done.
18. (Optional) Assign the indicator to any compulsory indicator group sets.
19. 点击**保存**。

Create or edit an indicator group¶

1. Open the Metadata Management app and click Indicators and Predictors > Indicator groups.
2. Click the + New button.
3. Type a Name.
4. 选择指标并分配它们。
5. 点击**保存**。

Create or edit an indicator group set¶

1. Open the Metadata Management app and click Indicators and Predictors > Indicator group sets.
2. Click the + New button.
3. Fill in the Name, Short name and Description fields — see Common metadata object fields.
4. If you want to use the indicator group set as a dimension in analytics, select Data dimension.
5. Select Compulsory if every indicator should be required to belong to a group in this group set.
6. 选择指标组并分配它们。
7. 点击**保存**。

Merge indicators¶

Merging consolidates two duplicate indicators into one, reassigning all references from the source to the target and then deleting the source.

Note

You must have the F_INDICATOR_MERGE authority to access this page.

1. Open the Metadata Management app and navigate to Indicators > Merge Indicators (or go directly to #/indicators/merge).
2. Search for and select the source indicator (the one to be removed).
3. Search for and select the target indicator (the one to keep).
4. Review the summary of changes.
5. Click Confirm merge.

Warning

Merging is irreversible. The source indicator will be permanently deleted.

Merge indicator types¶

Note

You must have the F_INDICATOR_TYPE_MERGE authority to access this page.

1. Open the Metadata Management app and navigate to Indicator types > Merge Indicator types (or go directly to #/indicatorTypes/merge).
2. Search for and select the source indicator type (the one to be removed).
3. Search for and select the target indicator type (the one to keep).
4. Review the summary of changes.
5. Click Confirm merge.

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions. Indicators and indicator types both support merging; see Merge.

Manage organisation units¶

在本节中，您将学习如何：

• Create a new organisation unit and build up the organisation unit hierarchy
• Create organisation unit groups, group sets, and assign organisation units to them
• 修改组织单位的层次结构

About organisation units¶

The organisation unit hierarchy defines the organisation structure of DHIS2, for example how health facilities, administrative areas and other geographical areas are arranged with respect to each other. It is the where dimension of DHIS2, similar to how periods represent the when dimension.

The organisation unit hierarchy is built up by parent-child relations. In DHIS2, each of these nodes is an organisation unit. A country might for example have eight provinces, and each province might have a number of districts as children. Normally, the lowest levels consist of facilities where data is collected. Data collecting facilities can also be located at higher levels, for example national or provincial hospitals. You can therefore create skewed organisation trees in DHIS2.

• You can only have one organisation hierarchy at a time.
• You can have any number of levels in a hierarchy. Typically national organisation hierarchies in public health have four to six levels.
• You can create additional classifications by using organisation unit groups and organisation unit group sets, for example to create parallel administrative boundaries to the health care sector.
• It is recommended to use organisation unit groups to create a non-geographical hierarchy.
• An organisation unit can only be a member of a single organisation unit group within an organisation unit group set.
• An organisation unit group can be part of multiple organisation unit group sets.
• The organisation unit hierarchy is the main vehicle for data aggregation on the geographical dimension.
• When you close an organisation unit, you cannot register or edit events to this organisation unit in the Capture app.

Important

You can change the organisation unit hierarchy after you have created it, even if organisation units are collecting data. However, DHIS2 always uses the latest hierarchy for data aggregation. So if you change the hierarchy, you lose the temporal representation of the hierarchy over time.

For example: District A is sub-divided into District B and District C. Facilities that belonged to District A are reassigned to District B and C. Any historical data entered before the split is still registered as belonging to District B and C, not to the obsolete District A.

In the Metadata Management app, you manage the following organisation unit objects:

Workflow¶

1. Create organisation units (representing the hierarchy nodes, top to bottom).
2. 创建组织单位组。
3. 创建组织单位组集。
4. Assign organisation units to groups.
5. Assign organisation unit groups to group sets.

Create or edit an organisation unit¶

1. Open the Metadata Management app and click Organisation units > Organisation units. On the left side of the screen, a tree displays the existing organisation unit hierarchy. Click on an organisation unit in the tree to select it; it will be highlighted. When you add a new organisation unit, it is created as a child of the currently selected organisation unit.

2. Click the + New button.

3. Fill in Name, Short name, Code, Description, Comment and contact details (Contact person, Address, Phone number, Email) — see Common metadata object fields. Only Name is required.
4. Set an Opening date — the date from which the organisation unit is active. Organisation units with a future opening date will not appear in analytics or data entry unless explicitly requested.
5. (Optional) Set a Closed date — the date from which the organisation unit is no longer active. See Close an organisation unit.
6. (Optional) Add Geometry — point or polygon coordinates for the organisation unit.
7. 点击**保存**。

Create or edit an organisation unit group¶

Organisation unit groups allow you to classify organisation units for analysis and reporting.

1. Open the Metadata Management app and click Organisation units > Organisation unit groups.
2. Click the + New button.
3. Fill in the Name, Short name and (optionally) Code and Description fields — see Common metadata object fields.
4. (Optional) Assign a Symbol for use on maps.
5. (Optional) Select a Color, used to display the group on maps and in group set reporting.
6. 选择组织单位并分配它们。
7. 点击**保存**。

Create or edit an organisation unit group set¶

Organisation unit group sets allow you to classify multiple organisation unit groups into a single dimension, for example "Type of facility" (comprising groups such as "Hospital", "Health centre" and "Community health post").

1. Open the Metadata Management app and click Organisation units > Organisation unit group sets.
2. Click the + New button.
3. Fill in the Name and Description fields — see Common metadata object fields.
4. Select Compulsory if you want every organisation unit to be required to be a member of a group in this group set.
5. Select Data dimension to make this group set available as a dimension in analytics.
6. 选择组织单位组并分配它们。
7. 点击**保存**。

Assign names to organisation unit levels¶

When you have created your organisation unit hierarchy, you can assign names to each level. For example, in a national hierarchy, Level 1 might be named "National", Level 2 "Province", Level 3 "District" and Level 4 "Facility".

1. Open the Metadata Management app and click Organisation units > Organisation unit levels.
2. For each level in the list, type a name.
3. 点击**保存**。

Move organisation units within a hierarchy¶

You can move organisation units within the hierarchy by changing the parent of a selected organisation unit. This is done on a dedicated page rather than in the main organisation unit list.

1. Open the Metadata Management app and click Organisation units > Hierarchy operations in the left-hand sidebar.
2. In the left-hand Organisation units to move tree, select the organisation unit or units you want to move. Hold Shift or Ctrl to select more than one.

Note

If a selected organisation unit has children, all of them move with it to the new parent.

1. In the right-hand Move into tree, select the organisation unit you want to move the selection to.
2. Click Move x organisation units, where x is the number of organisation units you selected.

Your changes are reflected in the hierarchy tree immediately.

Close an organisation unit¶

When you close an organisation unit, you cannot register or edit events for this organisation unit in the Capture app.

1. Open the Metadata Management app and click Organisation units > Organisation units.
2. In the object list, click the options menu and select Edit.
3. Enter a Closed date.
4. 点击**保存**。

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions. Organisation unit levels are the exception — they only support assigning names and translating (via the Translate icon directly, since they have no options menu); they don't support Show details, Sharing settings, Delete or Clone.

Manage validation rules¶

About validation rules¶

验证规则基于表达式。该表达式定义了一个 数据元素值之间的关系。该表达式形成一个 条件符合某些逻辑条件。

该表达式包括：

• 左侧

• 右边

• 操作员

验证规则断言所接种的疫苗总数 婴儿小于或等于婴儿总数。

左侧和右侧必须返回数字值。

In the Metadata Management app, you manage the following validation rule objects:

About sliding windows¶

You can use sliding windows to group data across multiple periods as opposed to selecting data for a single period. Sliding windows have a size, that is to say, the number of days to cover, a starting point and an end point. The example below shows disease surveillance data:

• 橙色部分中的数据，根据当前数据选择数据 期。有一个阈值，每周计算一次 或期间，并显示在“结果”部分中。

• 蓝色部分中的数据是滑动窗口。它选择数据 从过去7天开始。 “结果”显示了总数 确诊病例。

• 验证规则可确保在总数达到 案例数超出了该时间段的阈值。

See also: How to use sliding windows when you're Creating or editing a validation rule.

关于验证规则组¶

A validation rule group allows you to group related validation rules. When you run a validation rule analysis, you can choose to run all of the validation rules in your system, or just the validation rules in one group.

About validation notifications¶

You can configure a validation rule analysis to automatically send notifications about validation errors to selected user groups. These messages are called validation notifications. They are sent via the internal DHIS2 messaging system.

您可以将验证规则通知作为单独的消息或作为 消息摘要。这很有用，例如，如果您要发送 有关高优先级疾病暴发的个别消息和摘要 用于低优先级常规数据验证错误。

About validation rule functions¶

You can use the following functions in a validation rule left side or right side:

Create or edit a validation rule¶

1. Open the Metadata Management app and click Validations > Validation rules in the left-hand sidebar.

2. Click the + New button.

3. Type a Name.

   在验证规则中，名称必须唯一。

4. （可选）在 Code 字段中，分配一个代码。

5. （可选）输入**说明**。

6. (Optional) Type an Instruction. This is what users will see, both in validation analysis and when reviewing rule violations during data entry, so it's recommended to always provide an instruction that makes it clear what action a user should take when the rule is triggered.

7. 选择**重要性**：高，中**或**低。

8. Select a Period type.

9. 选择一个**操作员**：必填项，等于， 独家配对，大于，大于或等于 或**不等于**。

   **必填对**运算符允许要求该数据值 必须为表格的左侧和右侧输入表格 表情，还是双方都没有。这意味着您可以要求 如果填写表单中的一个字段，则一个或多个其他字段 还必须填写。

   **独占对**可以断言是否存在任何值 左侧，则右侧应没有值（或 反之亦然）。这意味着构成规则的数据元素 任何一方都应相互排斥，因为 给定的时间段/组织单位/属性选项组合。

10. 创建表达式的左侧：

    1. Click Left side.

    2. Select Sliding window if you want to view data relative to 您比较的时期。另请参见关于验证 rules.

    3. Select a Missing value strategy. This selection sets how the 如果数据丢失，系统将评估验证规则。

       选项	说明
       Skip if any value is missing	The validation rule will be skipped if any of the values which compose the expression are missing. This is the default option. Always select this option when you use the Exclusive pair or Compulsory pair operator.
       Skip if all values are missing	The validation rule will be skipped only if all of the operands which compose it are missing.
       Never skip	The validation rule will never be skipped in case of missing data, and all missing operands will be treated effectively as a zero.

    4. 输入**说明**。

    5. 根据可用的数据元素构建一个表达式， 项目对象，组织单位，计数和常量。

       在右窗格中，双击想要的数据对象。 包括在表达式中。与数学结合 运算符位于左窗格下方。

    6. 点击**保存**。

11. 创建表达式的右侧：

    1. Click Right side.

    2. Select a Missing value strategy. This selection sets how the 如果数据丢失，系统将评估验证规则。

       选项	说明
       Skip if any value is missing	The validation rule will be skipped if any of the values which compose the expression are missing. This is the default option. Always select this option when you use the Exclusive pair or Compulsory pair operator.
       Skip if all values are missing	The validation rule will be skipped only if all of the operands which compose it are missing.
       Never skip	The validation rule will never be skipped in case of missing data, and all missing operands will be treated effectively as a zero.

    3. Select Sliding window if you want to view data relative to 您比较的时期。另请参见关于验证 rules.

    4. 输入**说明**。

    5. 根据可用的数据元素构建一个表达式， 项目对象，组织单位，计数和常量。

       在右窗格中，双击想要的数据对象。 包括在表达式中。与数学结合 运算符位于左窗格下方。

    6. 点击**保存**。

12. (Optional) Choose which Organisation unit levels this rule 应该进行评估。将此留空将导致 要在所有级别上评估的验证规则。

13. (Optional) Select Skip this rule during form validation to stop this rule from being triggered during data entry.

14. 点击**保存**。

Create or edit a validation rule group¶

1. Open the Metadata Management app and click Validations > Validation rule groups in the left-hand sidebar.

2. Click the + New button.

3. Type a Name.

4. （可选）在 Code 字段中，分配一个代码。

5. （可选）输入**说明**。

6. Double-click the Validation rules you want to assign to the 组。

7. 点击**保存**。

Create or edit a validation notification¶

1. Open the Metadata Management app and click Validations > Validation notifications in the left-hand sidebar.

2. Click the + New button.

3. Type a Name.

4. （可选）在 Code 字段中，分配一个代码。

5. Select Validation rules.

6. Select Recipient user groups.

7. (Optional) Select Notify users in hierarchy only.

   如果选择此选项，系统将过滤收件人 用户。 （系统从收件人那里获取收件人用户 用户组。）过滤器基于 收件人用户所属。链接到组织单位的用户 是违规组织单位的祖先 发生后将收到验证通知。系统将 忽略其他用户，这些用户将不会收到验证 通知。

8. 创建消息模板：

   1. 创建**主题模板**。

      Double-click the parameters in the Template variables field 将它们添加到您的主题。

   2. 创建**消息模板**。

      Double-click the parameter names in the Template variables 字段以将其添加到您的消息中。

9. 点击**保存**。

For Show details, Translate, Delete and Clone, see Common actions. Validation rule groups additionally support Sharing settings (also in Common actions).

Manage attributes¶

About attributes¶

您可以使用元数据属性将其他信息添加到 元数据对象。除了每个标准属性 这些对象可能有助于存储其他信息 属性，例如数据元素的收集方法。

In the Metadata Management app, you manage the following attribute objects:

创建或编辑属性¶

1. Open the Metadata Management app and click Other > Attributes in the left-hand sidebar.

2. Click the + New button.

3. 在 名称 字段中，输入属性的名称。

   Each attribute must have a unique name.

4. （可选）在 Code 字段中，分配一个代码。

5. 选择**值类型**。

   如果为属性提供的值与该值不匹配 键入您将收到警告。

6. (Optional) Select an Option set to restrict the attribute's values to a predefined list.

7. (Optional) In the Sort order field, type a number to control where this attribute appears relative to other attributes when editing an object's form. Attributes without a sort order are shown afterwards, in alphabetical order.

8. 选择所需的选项，例如：

   • 选择**强制**如果希望对象始终具有 动态属性。

   • Select Unique if you want the system to enforce that values 对于特定的对象类型是唯一的。

9. 点击**保存**。

   动态属性现在可用于您分配的对象 它来。

For Show details, Translate, Delete and Clone, see Common actions. Attributes do not support Sharing settings.

Manage constants¶

About constants¶

常量是可以供用户使用的静态值 在数据元素和指标中。一些指标，例如“夫妇年 保护率”取决于通常不会转换的常数 时间。

In the Metadata Management app, you manage the following constant objects:

Create or edit a constant¶

1. Open the Metadata Management app and click Other > Constants in the left-hand sidebar.

2. Click the + New button.

3. 在 名称 字段中，输入常量的名称。

4. In the Short name field, type an abbreviated name of 常数。

5. （可选）在 Code 字段中，分配一个代码。

6. In the Description field, type a brief, informative description 常数。

7. 在 值 字段中，定义常量的值。

8. 点击**保存**。

   现在可以使用该常数。

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions.

Manage option sets¶

About option sets¶

选项集提供了一个预定义的下拉列表（枚举），用于 DHIS2。您可以定义任何类型的选项。

一个名为“投放类型”的选项集将具有以下选项：“普通”， “违反”，“剖腹产”和“协助”。

Create or edit an option set¶

重要

选项集必须具有代码和名称。您可以更改 名称，但是您无法更改代码。所有的名称和代码 选项即使在不同的选项集之间也必须是唯一的。

1. Open the Metadata Management app and click Option sets > Option sets in the left-hand sidebar.

2. Click the + New button.

3. In the Primary details tab, fill in Name, Code and (optionally) Description — see Common metadata object fields — select a Value type, then click Save.

4. 对于您需要的每个选项，执行以下任务：

   1. Click the Options tab.

   2. Click the + New button.

   3. Type a Name and a Code. Optionally also select a **颜色**和**图标**将用于此选项 数据捕获应用项目。

   4. 按名称，代码/值或手动对选项进行排序。

   5. 点击**保存**。

Create or edit an option group¶

You can group and classify options within an option set by using option groups. This way you can create a subset of options in an option set. The main purpose of this is to be able to filter huge option sets into smaller, related parts.

Options that are grouped can be hidden or shown together in the Capture app through program rules.

Note

You cannot change the Option set selected in an Option group once it has been created.

1. Open the Metadata Management app and click Option sets > Option groups in the left-hand sidebar.
2. Click the + New button.
3. Fill in Name, Short name and Code — see Common metadata object fields.
4. Select the Option set this group belongs to.
5. Once an Option set is selected, assign the Options you want to group.
6. 点击**保存**。

Create or edit an option group set¶

Option group sets allow you to categorise multiple option groups into a set. The main purpose of the option group set is to add more dimensionality to your captured data for analysis.

Note

You cannot change the Option set selected in an Option group set once it has been created.

1. Open the Metadata Management app and click Option sets > Option group sets in the left-hand sidebar.
2. Click the + New button.
3. Fill in Name, Code and Description — see Common metadata object fields.
4. Select the Option set this group set belongs to.
5. (Optional) Select Data dimension to make the group set available to analytics as another dimension, in addition to the standard dimensions of "Period" and "Organisation unit".

6. Select option groups and assign them. Available option groups are displayed in the left panel. Option groups that are currently members of the option group set are displayed in the right hand panel.

7. 点击**保存**。

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions.

Manage legends¶

About legends¶

You can create, edit, clone, share, delete, show details for and translate legends, to make the maps you set up for your users meaningful. You create the maps themselves in the Maps app.

注意

连续图例必须包含结束和开始的图例项 具有相同值，例如：0-50和50-80。不设置图例 像这样的项目：0-50和51-80。这将在您的图例中造成空白。

Create or edit a legend¶

注意

不允许在图例中留有空白。

不允许有重复的图例项。

1. Open the Metadata Management app and click Other > Legends in the left-hand sidebar.

2. Click the + New button.

3. In the Name field, type the legend name.

4. （可选）在 Code 字段中，分配一个代码。

5. 创建您想在图例中包含的图例项：

   1. 选择**起始值**和**结束值**。

   2. Select Number of legend items.

   3. 选择一种配色方案。

   4. 点击**创建图例项**。

   Tip

   Click the options menu to edit or delete a legend item.

6. （可选）添加更多图例项：

   1. Click the + New button.

   2. 输入名称，然后选择一个开始值，一个结束值和一种颜色。

   3. 点击**确定**。

7. （可选）更改色阶。

   1. Click the color scale to view a list of color scale options, and select one.

   2. To customise a color scale, click the + New button. In the **Edit 图例项**对话框，单击色阶按钮并手动选择 颜色，或输入您的颜色值。

8. 点击**保存**。

A finished legend with four items looks like this:

Assign a legend to an indicator or data element¶

You can assign a legend to an indicator or a data element in the Metadata Management app, either when you create the object or when you edit it. When you then select that indicator or data element in the Maps app, the system automatically selects the assigned legend.

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions.

See also¶

• Using the GIS app

Manage predictors¶

关于预测变量¶

预测变量定义如何从表达式生成聚合数据值 包含汇总和/或事件数据。预测值可以基于：

• 来自与预测值相同时期的数据，和/或

• 来自预测值之前期间的数据

Data from the same period¶

预测器可以使用与预测值相同周期的数据。 例如，您可以计算具有 通过使用预测器表达式，例如：数据元素的非零值：

  if（＃{ji7o0ILHuU2}！= 0，1，0） 

When you run this predictor at the organisation unit level where the data is collected, it will store 1 as the predicted value if the data element has a nonzero value for that organisation unit, otherwise 0. (If the data element that you predict into does not store zeros, then zeros will not be stored in the database, to save space.) You can then sum this predicted value in analytics at a higher organisation unit level, to count the number of organisation units with a nonzero value that are under each organisation unit in the report.

Data from previous periods¶

当您指定汇总时，预测变量将使用以前期间的数据 函数，例如sum（）或avg（）。例如，以下生成器表达式 标识一个值，该值是平均值加上标准偏差的两倍 前期数据：

  avg（＃{ji7o0ILHuU2}）+ 2 * stddev（＃{ji7o0ILHuU2}） 

Data from the same and previous periods¶

A predictor expression can access data both from same period as the prediction and previous periods by accessing data both within an aggregate function (for previous periods), and outside any aggregate function (for the same period). For example, an expression like the following can be used to take a balance of something from the previous period (#{KOh02hHko7C}), add to that the net change in this period (#{ji7o0ILHuU2}), resulting in the balance for this period:

  sum（＃{KOh02hHko7C}）+＃{ji7o0ILHuU2}  

The first data value is inside an aggregate expression (sum) to indicate that it is sampling previous period data (even if there is only one previous period), while the second data value is not in an aggregate function to indicate that it is referencing data from the same period.

如果需要，可以在同一引用中引用预测变量的输出数据元素 预测变量的表达。例如，此示例中的表达式可以 用于预测一段时间内的余额，然后将该值与 下一个期间的变化以计算下一个期间的余额。 当预测变量跨多个期间运行时，将在以下期间处理这些期间 按时间顺序排列，可以使用较早时期的结果 作为以后的输入。

Predictor organisation unit levels¶

您需要为预测变量的输出选择一个或多个组织单位级别。 预测变量生成的所有值都存储在组织单位的 您选择的级别。预测变量表达式中的每个项目都是值的总和 为该组织单位存储的信息（如果有的话）以及组织中存储的所有值 低于该单位（如果有）的单位。

注意

在配置预测变量时，必须选择一个或多个组织单位级别 将输出预测数据。如果未选择任何级别，则不会预测 将生成值。

警告

如果您想在分析报告中使用预测值或 其他预测，请勿选择多个组织单位级别。 当您选择多个级别时，较高级别的预测将 还包括在较低级别的预测中使用的任何数据。如果预测 随后从多个级别将其用于分析或表达式中 其他预测变量，这可能会导致重复计算，因为 较高级别的值包括较低级别的预测值。

You may select multiple organisation unit levels if you use the predicted values only in validation rules. For example in disease surveillance, you could have a validation rule alert if an actual value is higher than the range of expected values for that period based on previous period data. To do this, you could create a predictor to compute the average plus twice the standard deviation of previous period data. You could use a validation rule to compare this highest expected value with the actual value. You could run the predictor and validation rule at multiple levels to detect different outbreak scenarios. In one scenario, there might be a significant increase in one facility that exceeds its expected range, but the district containing that facility might not exceed its expected range because the district values are combined with many other facilities. Yet in another scenario, there may be a moderate increase in several facilities that does not exceed the expected range for each facility (because the standard deviation for each facility may be high), but it does exceed the expected range for the district (because the standard deviation for the district as a whole may be lower).

If you want to generate predictions at multiple levels, you could also use different predictors at different levels. For example, you might want to be alerted if the value at one level exceeds the average plus twice the standard deviation, but alerted at another level if it exceeds the average plus 1.8 times the standard deviation. If you want, you could configure the different predictors to have the same output data element. If you use the same output data element, this will still work with validation rules at different organisation unit levels, but you also must be careful not to use the results in analytics or in other predictor calculations to avoid double counting.

In the Metadata Management app, you manage the following predictor objects:

抽样前期¶

预测变量可以生成过去一段时间的数据值， 现在或将来。这些值基于预测期间的数据， 和/或来自预测时期之前的时期的采样数据。

如果您只需要进行预测的同一时期的数据， 那么您无需阅读本节。本节介绍如何 从预测周期之前的周期中采样数据。

Three fields control which previous periods are sampled — Sequential sample count, Annual sample count and Sequential skip count — and a fourth, the Sample skip test, can exclude individual periods that would otherwise be included. They appear together in the predictor's Logic section:

The rest of this section explains what each one does. In the diagrams, the period being predicted is marked in one colour and the periods that get sampled in another.

Sequential sample count¶

A predictor's Sequential sample count gives the number of immediate previous periods to sample. For example, if a predictor's period type is Weekly and the Sequential sample count is 4, this means to sample four previous weeks immediately preceding the predicted value week. So the predicted value for week 9 would use samples from weeks 5, 6, 7, and 8:

If a predictor's period type is Monthly and the Sequential sample count is 4, this means to sample four previous months immediately preceding the predicted value month. So the predicted value for May would use samples from January, February, March and April:

The Sequential sample count can be greater than the number of periods in a year. For example, if you want to sample the 24 months immediately preceding the predicted value month, set the Sequential sample count to 24:

Sequential skip count¶

A predictor's Sequential skip count tells how many periods should be skipped immediately previous to the predicted value period, within the Sequential sample count. This could be used, for instance, in outbreak detection to skip one or more immediately preceding samples that might in fact contain values from the beginning of an outbreak that you are trying to detect.

For example, if the Sequential sample count is 4, but the Sequential skip count is 2, then the two samples immediately preceding the predicted period will be skipped, resulting in only two periods being sampled:

Annual sample count¶

A predictor's Annual sample count gives the number of previous years for which samples should be collected at the same time of year. This could be used, for instance, for disease surveillance in cases where the expected incidence of the disease varies during the year and can best be compared with the same relative period in previous years. For example, if the Annual sample count is 2 (and the Sequential sample count is zero), then samples would be collected from periods in the immediately preceding two years, at the same time of year.

Sequential and annual sample counts together¶

You can use the sequential and annual sample counts together to collect samples from a number of sequential periods over a number of past years. When you do this, samples will be collected in previous years during the period at the same time of year as the predicted value period, and also in previous years both before and after the same time of year, as determined by the Sequential sample count number.

For example, if the Sequential sample count is 4 and the Annual sample count is 2, samples will be collected from the 4 periods immediately preceding the predicted value period. In addition samples will be collected in the previous 2 years for the corresponding period, as well as 4 periods on either side:

Sequential, annual, and skip sample counts together¶

You can use the Sequential skip count together with the sequential and annual sample counts. When you do this, the Sequential skip count tells how many periods to skip in the same year as the predicted value period. For example, if the Sequential sample count is 4 and the Sequential skip count is 2, then the two periods immediately preceding the predicted value period will be skipped, but the two periods before that will be sampled:

If the Sequential skip count is equal to or greater than the Sequential sample count, then no samples will be collected for the year containing the predicted value period; only periods from past years will be sampled:

Sample skip test¶

您可以使用* Sample skip test *（样本跳过测试）来跳过某些时段的样本 根据测试结果， 在那些时期内表达。例如，可以在 疾病暴发检测，样本跳过测试可以识别 以前的疾病暴发，将那些样本从预测中排除 非爆发基准期望值。

• Sample skip test *是一个应返回值的表达式 是或否，表示是否应跳过该时间段。 它可以是一个测试上一时期中任何数据值的表达式。 例如，它可以测试是否有明确输入的数据值 表示应跳过上一个期间。或者可以 将一段时间内的先前预测值与实际值进行比较 记录该时间段，以确定是否应跳过该时间段。

*样本跳过测试*为*真*的任何时期都不会 采样。例如：

Predictors and category option combinations (disaggregations)¶

The category option combination (disaggregation) for predictor output data is chosen in one of three ways:

1. Default category option combination.

If the predictor's output data element has no disaggregations (category combination "None", also known as the default category combination), then all predictor output data will be made in the default category option combination. In this case, predictor output data is not disaggregated.

1. Fixed category option combination

If the predictor's output data element has a category combination other than "None", you can choose a fixed disaggregation for the predictor "Output category option combo". If you do so, all output from this predictor will have this category option combination.

For example, if the output data element has a category combination of "Sex and Age", you can decide that all predictor output will go to a category option combination such as "Female under 5", "Male 5 - 10", or any other.

1. Use the input category option combo (available in v40.1 and following)

If the predictor's output data element has a category combination other than "None", you can choose "Predict according to input category option combo" as the value for "Output category option combo". If you do so, a different prediction is made for each category option combo in the output data element's category combination (that is, one prediction for "Female under 5", a second for "Male under 5", a third for "Female 5 - 10", and so on).

(In v40.1, this feature is enabled by selecting the choice "\<no value>" for "Output category option combo".)

> **Tip**
>
> In some installations, the options in a category may have changed over time.
For example, a category "Age" may have had three options "1-5", "6-10", and "over 10",
but then was changed to only have two options "10 and under" and "over 10".
Historical data may have values with any of these options.
To use a predictor with such data, you can define another category
such as "Reporting Age" with all the options ever used, such as "1-5", "6-10",
"10 and under", and "over 10". Use this category in the category combination
for the predictor's output data element. This means that all the input category
options will be reflected in the output.
>
> If you then want a single report for the output data that covers
all the disaggregations, you can use category option groups
in a category option group set. For example use a category option group
"10 and under" that contains category options "1-5", "6-10", and "10 and under".

Predictors and attribute option combinations¶

If the input data to a prediction has attribute option combinations, a different prediction will be made for each attribute option combination where there is data.

For example, you could use attribute option combinations to represent different projects on your system. The predictor will generate a value with the attribute option combination for Project A when it finds input data for Project A; it will generate a value with the attribute option combination for Project B when it finds input data for Project B; and so on.

For any input data without attribute option combinations (in other words, with the default attribute option combination), predictions are generated using the default attribute option combination. If you don't use attribute option combinations in the data, they will not be used in predictor output data.

If you use attribute option combinations and also "Predict according to input category option combo", there will be a separate prediction for each combination of disaggregation and attribute option combination. For example, there could be a prediction for Project A data for "Female under 5", a prediction for Project A data for "Male under 5", a prediction for Project B data for "Female under 5", and so on.

Create or edit a predictor¶

1. Open the Metadata Management app and click Indicators and Predictors > Predictors in the left-hand sidebar.

2. Click the + New button.

3. In the Name field, type the predictor name.

4. In the Short name field, type a short name for the predictor.

5. （可选）在 Code 字段中，分配一个代码。

6. （可选）输入**说明**。

1. Select an Output data element. Values generated by this 预测变量存储为与此数据关联的聚合数据 元素和预测周期。

   该值根据数据的值类型四舍五入 元素：如果值类型是整数类型，则预测值 舍入到最接近的整数。对于所有其他值类型， 数字四舍五入到四位有效数字。 （但是，如果有 小数点左边多于四位数，它们不是 替换为零。）

2. (Optional) Select an Output category option combo. This dropdown will only show if the selected data element has a category combination other than "None". If so, you can select which disaggregation category option combo you would like to output to, or you can select "Predict according to input category option combo" (see the discussion above).

3. Select a Period type.

4. Assign one or more organisation unit levels. The output value will be 分配给此级别（或这些级别）的组织单位。 For aggregate data, the input values depend on the Organisation units providing data setting described below. For tracker-based data, the input values come from the organisation unit to which the 分配了输出，或从输出下方的任何较低级别进行分配 组织单位。

5. Organisation units providing data controls where the input values come from for aggregate data. If "at selected levels only", only organisation units at the selected levels are included. If "At selected levels and all levels below" is selected, organisation units at the selected level(s) and all organisation units below are also included.

6. 创建一个**生成器**。生成器是用于计算预测值的表达式。

   

   1. Type a Description of the generator expression.

   2. Select a Missing value strategy. This sets whether the system evaluates the generator expression at all when some of the data it references is missing.

      选项	说明
      Skip if any value is missing	The expression is not evaluated if any of the values which compose it are missing, so no value is predicted. This is the default option.
      Skip if all values are missing	The expression is not evaluated only if every value which composes it is missing. If at least one value is present, the expression runs and the missing values are treated as zero.
      Never skip	The expression is always evaluated, and every missing value is treated as zero.

      The strategy decides whether the expression runs. That is separate from what the expression does with a missing value once it is running:

      Situation	Skip if any value is missing	Skip if all values are missing	Never skip
      The only item referenced is missing	No value predicted	No value predicted	Treated as 0
      One item missing, others have values	No value predicted	Missing item treated as 0	Missing item treated as 0
      isNull( item ) where item is missing	Never evaluated	true	true
      isNotNull( item ) where item is missing	Never evaluated	false	false
      firstNonNull( item, fallback ) where item is missing	Never evaluated	Returns fallback	Returns fallback

      Note

      A missing value and a zero are not the same thing. DHIS2 does not store zeros for a data element unless Store zero data values is switched on for it, so a field that looks like it holds zero may hold no value at all. If your expression needs to tell "reported zero" apart from "not reported", test with isNull() and use Skip if all values are missing or Never skip — under Skip if any value is missing the expression never runs, so the test can never fire.

   3. 输入生成器表达式。您可以通过以下方式构建表达式 selecting data elements for aggregate data, program data elements, attributes or indicators, organisation unit counts, constants or reporting rates.

      要使用过去的采样数据，应将选择的所有项目包含在以下聚合函数之一中（请注意，这些函数名区分大小写）：

      Aggregate function	Means
      avg(x)	Average (mean) value of x
      count(x)	Count of the values of x
      max(x)	Maximum value of x
      median(x)	Median value of x
      min(x)	Minimum value of x
      percentileCont(p, x)	Continuous percentile of x, where p is the percentile as a floating point number between 0 and 1. For example, p = 0 will return the lowest value, p = 0.5 will return the median, p = 0.75 will return the 75th percentile, p = 1 will return the highest value, etc. Continuous means that the value will be interpolated if necessary. For example, percentileCont( 0.5, #{FTRrcoaog83} ) will return 2.5 if the sampled values of data element FTRrcoaog83 are 1, 2, 3, and 4.
      stddev(x)	Standard deviation of x. This function is equivalent to stddevSamp. It's suggested that you use the function stddevSamp instead for greater clarity.
      stddevPop(x)	Population standard deviation of x: sqrt( sum( (x - avg(x))^2 ) / n )
      stddevSamp(x)	x 的样本标准差：sqrt( sum( (x - avg(x))^2 ) / ( n - 1 ) )。请注意，当只有一个样本时不会计算此值。
      sum(x)	Sum of the values of x

      Note

      Any items inside an aggregate function will be evaluated for all sampled past periods, and then combined according to the formula inside the aggregate function. Any items outside an aggregate function will be evaluated for the period in which the prediction is being made.

      您可以通过单击（或 输入）表达式字段下方的任何元素： （） * / +-天数。 可以通过键入常数来添加常数。那些日子 选项将 [days ]插入到解析为 数据来源期间的天数。

      您还可以在您的计算机中使用以下非聚合函数 表达式，在聚合函数内部或包含 聚合函数，或独立于聚合函数：

      Function	Means
      contains(expr, sub1, ...)	Searches an expression for one or more substrings. Returns true if the expression contains all the substrings. For example, the following are all true: contains("abcd", "abcd"); contains("abcd", "b"); and contains("abcd", "ab", "bc"). Comparisons are case-sensitive.
      containsItems(expr, item1, ...)	Searches an expression for one or more items. The expression is made up of comma-separated elements. containsItems returns true if every item exactly matches an element in the expression. For example, containsItems("abcd", "abcd") and containsItems("ab,cd", "ab", "cd") are true, but containsItems("abcd", "b") and containsItems("abcd", "ab", "bc") are false. Comparisons are case-sensitive. containsItems can be used for multi-valued data elements to see if an item is contained in the data element values.
      if(test, valueIfTrue, valueIfFalse)	Evaluates test which is an expression that evaluates to a boolean value -- see Boolean expression notes below. If the test is true, returns the valueIfTrue expression. If it is false, returns the valueIfFalse expression.
      is(expr1 in expression [, expression ...])	Returns true if expr1 is equal to any of the following expressions, otherwise false.
      isNull(item)	Returns the boolean value true if the item is null (missing), otherwise returns false. The item can be any selected item from the right (data element, program data element, etc.).
      isNotNull(item)	Returns true if the item value is not missing (not null), otherwise false.
      firstNonNull(item [, item ...])	Returns the value of the first item that is not missing (not null). Can be provided any number of arguments. Any argument may also be a numeric or string literal, which will be returned if all the previous items have missing values.
      greatest(expression [, expression ...])	Returns the greatest (highest) value of the expressions given. Can be provided any number of arguments.
      least(expression [, expression ...])	Returns the least (lowest) value of the expressions given. Can be provided any number of arguments.
      log(expression [, base ])	Returns the natural logarithm (base e) of the numeric expression. If an integer is given as a second argument, returns the logarithm using that base.
      log10(expression)	Returns the common logarithm (base 10) of the numeric expression.
      normDistCum(x [,mean [,stddev]])	Returns the cumulative distribution function (CDF) value for x given the normalized distribution described by the mean and stddev. Equivalent to Excel NORM.DIST(x,mean,stddev,TRUE) or LibreOffice NORMDIST(x,mean,stddev,1). If stddev is not given, it is computed from past sampled values of x. If neither mean nor stddev are given, they are computed from past sampled values of x. See examples.
      normDistDen(x [,mean [,stddev]])	Returns the probability density function (PDF) value for x given the normalized distribution described by the mean and stddev. Equivalent to Excel NORM.DIST(x,mean,stddev,FALSE) or LibreOffice NORMDIST(x,mean,stddev,0). If stddev is not given, it is computed from past sampled values of x. If neither mean nor stddev are given, they are computed from past sampled values of x. See examples.
      null	Returns no result. For example, if( #{FH8ab5Rog83}<0, null, 1 ) returns nothing if the data element value is less than 0, otherwise 1.
      orgUnit.ancestor(orgUnitUid [, orgUnitUid ...])	Returns true if the organisation unit is a descendant of any of the (1 or more) organisation units, otherwise false.
      orgUnit.dataSet(dataSetUid [, dataSetUid ...])	Returns true if the organisation unit is assigned to any of the (1 or more) data sets, otherwise false.
      orgUnit.group(ouGroupUid [, ouGroupUid ...])	Returns true if the organisation unit is a member of any of the (1 or more) organisation unit groups, otherwise false.
      orgUnit.program(programUid [, programUid ...])	Returns true if the organisation unit is assigned to any of the (1 or more) programs, otherwise false.
      removeZeros(expression)	Returns nothing if the expression value is 0, otherwise returns the expression value.
      .maxDate(yyyy-mm-dd)	For a data element (not program data), value from periods ending on or before a maximum date.
      .minDate(yyyy-mm-dd)	For a data element (not program data), value from periods starting on or after a minimum date.

      布尔表达式注释：**布尔表达式必须求值 to **true or false. The following operators may be used to 比较两个值，它们得出一个布尔表达式： <, >， \！=，==，> =和 <=。以下运算符可用于 combine two boolean expressions: && or the keyword and (logical and), and || or the keyword or (logical or). The unary operator ! or the keyword not may be used to negate a 布尔表达式。

      Generator expression examples:

      Generator expression	Means
      sum(#{FTRrcoaog83.tMwM3ZBd7BN})	Sum of the sampled values of data element FTRrcoaog83 and category option combination (disaggregation) tMwM3ZBd7BN
      avg(I{GSae40Fyppf}) + 2 * stddevSamp(I{GSae40Fyppf})	Average of the sampled values of program indicator GSae40Fyppf plus twice its sample standard deviation
      sum(D{IpHINAT79UW.eMyVanycQSC}) / sum([days])	Sum of all sampled values of data element eMyVanycQSC from program IpHINAT79UW divided by the number of days in all sample periods (resulting in the overall average daily value)
      sum(#{FTRrcoaog83}) + #{T7OyqQpUpNd}	Sum of all sampled values of data element FTRrcoaog83 plus the value of data element T7OyqQpUpNd in the period being predicted for (includes all disaggregations)
      1.2 * A{IpHINAT79UW.RKLKz1H20EE}	1.2 times the value of attribute RKLKz1H20EE of program IpHINAT79UW, in the period being predicted for
      if(isNull(#{T7OyqQpUpNd}), 0, 1)	If the data element T7OyqQpUpNd is null in the period being predicted, then 0, otherwise 1
      if(is(#{jeiTh8ahyae} in 'NEGATIVE','UNKNOWN'), 0, 1)	If the data element jeiTh8ahyae has value 'NEGATIVE' or 'UNKNOWN' then 0, otherwise 1
      percentileCont(0.5, #{T7OyqQpUpNd})	Continuous 50th percentile of the sampled values for data element T7OyqQpUpNd. Note that this is the same as median(#{T7OyqQpUpNd})
      if(count(#{T7OyqQpUpNd}) == 1, 0, stddevSamp(#{T7OyqQpUpNd}))	If there is one sample value present for data element T7OyqQpUpNd, then 0, otherwise the sample standard deviation of these sample values. (Note that if no samples are present then the stddevSamp returns no value, so no value is predicted.)
      normDistCum(#{T7OyqQpUpNd})	The cumulative distribution function for the current period value of data element T7OyqQpUpNd based on the normalized distribution defined by the mean and standard deviation of past sampled periods of data element T7OyqQpUpNd
      normDistCum( #{T7OyqQpUpNd}, median(#{T7OyqQpUpNd}) )	The cumulative distribution function for the current period value of data element T7OyqQpUpNd based on the distribution defined by the median (instead of the mean) and standard deviation of past sampled periods of data element T7OyqQpUpNd
      normDistCum( #{T7OyqQpUpNd}, avg(#{T7OyqQpUpNd}), stddev(#{T7OyqQpUpNd}) )	Same as normDistCum( #{T7OyqQpUpNd} )
      normDistDen( #{T7OyqQpUpNd}, avg(#{IKahz1Quie3}), stddev(#{IKahz1Quie3}) )	The probability density function for the current period value of data element T7OyqQpUpNd based on the distribution defined by the mean and standard deviation of past sampled periods of the different data element IKahz1Quie3
      normDistDen( median(#{T7OyqQpUpNd}), avg(#{IKahz1Quie3}), stddev(#{IKahz1Quie3}) )	The probability density function for the median of past sampled values of data element T7OyqQpUpNd based on the distribution defined by the mean and standard deviation of past sampled periods of the different data element IKahz1Quie3
      #{T7OyqQpUpNd}.minDate(2022-10-1)	Value on or after 1-Oct-2022
      #{T7OyqQpUpNd}.maxDate(2022-12-31)	Value on or before 31-Dec-2022
      #{T7OyqQpUpNd}.minDate(2022-10-1).maxDate(2022-12-31)	Value between 1-Oct-2022 and 31-Dec-2022

7. (Optional) Create a Sample skip test. The sample skip test tells 从样本中排除哪个先前时段（如果有）。

   1. Type a Description of the skip test.

   2. 输入样本跳过测试表达式。您可以建立 expression from the same items as the generator expression. As with the generator, you may click on (or type) any of the elements below the expression field: ( ) * / + - Days.

      The non-aggregating functions described above for generator expressions may also be used 在跳过测试中。

      该表达式的计算结果必须为布尔值** true 或 **false. See Boolean expression notes above.

      跳过测试表达式示例：

      Skip test expression	Means
      #{FTRrcoaog83} > #{M62VHgYT2n0}	The value of data element FTRrcoaog83 (sum of all disaggregations) is greater than the value of data element M62VHgYT2n0 (sum of all disaggregations)
      #{uF1DLnZNlWe} > 0	The value of data element uF1DLnZNlWe (sum of all disaggregations) is greater than zero
      #{FTRrcoaog83} > #{M62VHgYT2n0} || #{uF1DLnZNlWe} > 0	The value of data element FTRrcoaog83 (sum of all disaggregations) is greater than the value of data element M62VHgYT2n0 (sum of all disaggregations) or the value of data element uF1DLnZNlWe (sum of all disaggregations) is greater than zero

8. Enter a Sequential sample count value.

   这是计算应该进行多少个连续周期 及时返回样本数据进行计算。

9. Enter an Annual sample count value.

   这是计算应该追溯到多少年 用于计算的样本数据。

10. (Optional) Enter a Sequential skip count value.

    这是紧接在 采样值之前，应跳过预测值周期。

11. 点击**保存**。

Predictions by Data Element Group¶

You can use a single predictor to operate on all the data elements in a group instead of a different predictor for each data element. This can be used, for example, in logistics management when a data element is used for each commodity and a category option combination is used for each count related to that commodity.

The syntax is:

forEach ?de in :DEG:degUid --> main expression

where:

The predictor will execute once for each data element in the data element group. For each data element, instances of the variable in the main expression are replaced by that data element. The same data element is also used as the predictor output data element. The predicted value will be written to that data element using the configured output category option combination.

The predictor must be configured with an output data element, but it is effectively ignored when the predictor is run. It is suggested that you configure the predictor with one of the data elements in the data element group that the predictor will use. That way you can select a valid output category option combination for that data element.

At the time the predictor is created, the data element group must contain at least one data element of the type that you will use. (The data type of the data element is used during syntax checking of the predictor.)

Example 1¶

You have data elements that represent various commodities, all belonging to a data element group with UID aIMu0nieph7.

You have category option combinations with the following UIDs:

The following predictor generator expression will compute the stock balance at the beginning of the next period as ( starting balance + restock - used - lost ):

forEach ?de in :DEG:aIMu0nieph7 -->
sum( #{?de.Gvoecom5muL} + #{?de.CWa6eew5uco} - #{?de.nthohhie8Ba} - #{?de.Faey8Iphooy} )

The predictor configuration includes:

The predictor will execute once for each data element in the data element group. Because the aggregation function sum() is used in the predictor generator expression, all the values in the expression will be fetched from the previous period (since the sequential sample count is 1). The predictor will write out the starting balance for each data element for the periods within the predictor run start and end date, for organisation units at the selected level(s).

Predictions are always made forward through time. The starting balance predicted for one period can be used as an input to compute the starting balance of the following period.

Example 2¶

If you want to make predictions for the same period as the input data, just omit the aggregation function such as sum(). Adding to the previous example, say you have another category option combination that computes the net inventory change during the period:

You can use the following expression to compute the inventory change as ( restock - used - lost ):

forEach ?de in :DEG:aIMu0nieph7 -->
#{?de.CWa6eew5uco} - #{?de.nthohhie8Ba} - #{?de.Faey8Iphooy}

The output category option combo is:

Since there is no aggregation function such as sum() around the expression elements, the input data is taken from the same period as the predictor output.

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions.

Create or edit a predictor group¶

1. Open the Metadata Management app and click Indicators and Predictors > Predictor groups in the left-hand sidebar.

2. Click the + New button.

3. Type a Name. This field needs to be unique.

4. (Optional) In the Code field, assign a code. This field needs to 独一无二。

5. （可选）输入**说明**。

6. Double-click the Predictors you want to assign to the group.

7. 点击**保存**。

For Show details, Sharing settings, Translate, Delete and Clone, see Common actions. Predictor groups support the same actions as individual predictors, including Run now.

Manage push reports¶

Note

Push reports are not available in the Metadata Management app. If your DHIS2 instance still uses push reports, configure them in the legacy Maintenance app — see Manage push reports.

Manage external map layers¶

Note

Managing external map layers is planned to move into the Maps app and will not be replicated in the Metadata Management app. External map layers can still be configured in the legacy Maintenance app — see Manage external map layers.

Manage SQL views¶

DHIS2的SQL View功能将存储SQL视图定义 内部，然后在需要时实例化视图。

数据库管理员在创建数据库视图时必须小心 直接在DHIS2数据库中。例如，当资源表 生成，所有这些都将首先被删除然后重新创建。如果 任何SQL视图都依赖于这些表，完整性违规异常 将被抛出并且该过程将被中止。

SQL视图基于它们的反向字母顺序删除 DHIS2中的名称，并按常规字母顺序创建。这允许 您要在SQL视图之间具有依赖关系，因为视图仅取决于 其他按字母顺序排列的观点。对于 例如，“ ViewB”可以安全地依赖于“ ViewA”。否则，有意见 取决于其他视图会导致完整性违规错误。

创建一个新的SQL视图¶

To create a new SQL view, open the Metadata Management app and click Other > SQL views in the left-hand sidebar, then click the + New button.

The "Name" attribute of the SQL view will be used to determine the name of the table that DHIS2 will create when the view is materialized by the user. The "Description" attribute allows one to provide some descriptive text about what the SQL view actually does.

The "SQL type" attribute allows the creation of three kinds of views: - A "View" is stored in the database and regenerated when queried - A "Materialized View" is stored in the database and its results are cached in the database - A "Query" is not stored in the database

Finally, the "SQL query" should contain the SQL view definition.

Only SQL "SELECT" statements are allowed and certain sensitive tables (i.e., user information) are not accessible.

Press "Save" to store the SQL view definition. If you created a "View" or a "Materialized View", you must also run Run query to finish the creation of the SQL view.

Keep in mind that the columns returned by the used SELECT statement become table columns, that means they must be of a valid table column type. When functions are used it might be necessary to explicitly cast the result to a type by adding ::{TYPE} after the function.

For example, instead of jsonb_each (which would return a record type that cannot be a column type) use jsonb_each_text and cast the result to text, like in the below sample:

select jsonb_each_text(eventdatavalues)::text from ...

SQL views that call other SQL views¶

If you wish to make a SQL view that can be called be other SQL views, then its SQL type must be either "View" or a "Materialized View" (not "Query"). It also must have Run query run on it before being called.

For instance, if you created a view named Data element count with SQL type "View" and this SQL:

select count(*) as count from dataelement;

...then you could run Run query from the options menu and create a second SQL view named More than 100 data elements with this SQL:

select case when count > 100 then 1 else 0 end as result from _view_data_element_count;

SQL视图管理¶

In order to utilize the SQL views, simply click the view and from the options menu, choose Run query. Once the process is completed, you will be informed that a table has been created. The name of the table will be provided, and is composed from the "Description" attribute provided in the SQL view definition. Once the view has been generated, you can view it by clicking the view again, and selecting View results.

For Clone, see Common actions.

Tip

If you have a view which depends on another view, be careful about how the views are named. When analytics is run on the DHIS2 server, all views are dropped and then recreated: they are dropped in reverse alphabetical order and recreated in regular alphabetical order. So a view must come after the view it depends on alphabetically — for example "ViewB" can safely depend on "ViewA". If you name them the other way round, analytics may fail because the dependency will not exist yet when the dependent view is recreated.

Manage analytics table hooks¶

The Analytics Table Hooks functionality of DHIS2 stores SQL code that is run during different phases of the analytics table generation process.

See also /api/analyticsTableHooks in the Developer documentation.

Creating a new analytics table hook¶

To create a new analytics table hook, open the Metadata Management app and click Other > Analytics table hooks in the left-hand sidebar, then click the + New button.

Press "Save" to store the analytics table hook.

Manage Locales¶

可以在DHIS2中创建自定义语言环境。除了 系统提供的语言环境，您可能需要添加自定义 系统的语言环境（例如“英语”和“赞比亚”）。这将允许 您可以将元数据对象翻译成本地语言，或解决 使用共同元数据的国家之间的细微差异 定义。

The locale is composed of a language along with a country. Open the Metadata Management app and click Other > Locales in the left-hand sidebar, then click the + New button. Select the desired values and click Create Locale. This custom locale will now be available as one of the translation locales in the system.

Note

A locale cannot be edited once created.

Manage group membership¶

The legacy Maintenance app had a separate Metadata group editor screen for adding objects to groups in bulk. The Metadata Management app does not have an equivalent screen, and one is not planned. Group membership is instead edited from whichever side of the relationship you happen to be working on, using the standard transfer list:

• One object into many groups — open the object's own edit screen and assign it to groups there. For example, the Data element groups field in Create or edit a data element puts one data element into as many data element groups as you need, without leaving the data element.
• Many objects into one group — open the group's edit screen and assign the members there. For example, Create or edit a data element group assigns any number of data elements to a single group in one pass.

The same pattern applies to the other grouped object types in this chapter: category option groups, indicator groups, organisation unit groups, option groups and predictor groups, and their corresponding group sets.

To see or change which objects are currently in a group, open the group's list from the left-hand sidebar — for example Data elements > Data element groups — and edit the group you want.

Tip

When you are assigning members to a group, use the Show only unassigned items checkbox above the Available list, where it is offered, to hide objects that already belong to another group of the same type. See Using a transfer list component.