Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 65 Current »

Create New Metric

  1. On the main menu, click Perf, and then open the required project and open the Delivery Metrics page.

  2. Click Add → Metrics on the Delivery Metrics. The Metric Catalog is opened. 

  1. Find Custom Metric category in the left navigation menu and:

    • click (Create New Metric) button in the menu

                OR

    • open Custom Metric category and click Create New Metric.

Step 1: General

  1. Provide a Metric Name. The following requirements should be met:

  • The minimum length of the name should be 5 characters.

  • The name must be unique among all metrics, whether they are published or in draft status.

  • Cyrillic symbols are not permitted; only English characters are allowed.

  1. You can add a metric overview in the "Description." This text will be displayed in the metric catalog and in the widget tooltip.

  2. Select a Primary Dimension where you think the metric makes the biggest contribution.

  3. Optionally you can select Secondary Dimension(s) if the metric contributes in different areas.

  4. Click Next.

As soon as 'Next' is clicked and navigation is made to the next step, the metric is saved and becomes available in the metric catalog in draft state.

Custom metrics stepper.png

Step 2: Dataset selection

To build a custom metric, it is necessary to select a dataset based on which the metric will be created. A list of available datasets is presented in the left menu (see PERF Datasets). PerfQL provides powerful techniques for data manipulation, allowing you to create advanced custom metrics using formula-like expressions. You can achieve this by selecting, joining, filtering, and aggregating your data.

  1. To generate a dataset for a metric, write PerfQL statements in the PerfQL input field.

PerfQL Rules

  1. Only Select statements.

  2. Only one root statement per query.

  3. Only specified list of functions could be used (list of supported functions).

  4. Only specified list of datasets (tables and fields) could be used (list of allowed datasets).  

  5. See examples of PerfQL Queries for different cases

  1. To preview data based on your PerfQL request, click the Run Preview button, and you'll see the result table in the Result Preview area.

The result preview is read-only and displays only the first 100 records, but the metric will be calculated for all available data.

To save unfinished PerfQL work for later, click the 'Save Changes' button.

Step 3: Chart Preview

This step is used to create charts, which provide you with an efficient and simple way of visualizing data.

As a default setting, the first column from the Custom Selection (step 2) is a data for Axis X placement.

Subsequent columns are assigned to Axis Y.

By default, 3 columns are automatically included on Axis Y.

  1. To define the scope of data on Axis Y:

  • To add columns, click Add +.

  • To remove any column, click Bin.

  1. Select the chart type from the drop down menu (see Available Chart Types for PERF Custom Metrics). 

  2. To configure the series, click on the series name.

  3. In the opened menu, you can select colors, switch on values on the chart, add trend lines, and set thresholds.

Note

PERF allows you to set up different thresholds for each series on the chart.

You can view the thresholds on the chart if swithch off other series by clicking on the series names in the legend.

Rounding values on the chart

The Value Rounding feature empowers users to refine the precision of displayed numbers on the result chart, allowing them to customize the visual representation of data according to their preferences.

To enable value rounding on the chart labels:

  1. Select nessesary serias on the step 3.

  2. Check the checkbox labeled 'Show values on the chart.'

  3. Specify the number of decimal places for values on the result chart.

Please note that the rounding exclusively affects the labels view. If you wish to apply rounding to results in tooltips and drill-downs, adjust the rounding precision in the resulting dataset using the ROUND function.

Example of Rounding via PerfQL:

Select key, ROUND(avg(story_points), 2) as average_story_points
from ticket
group by sprint;

Step 4: Drill Downs

Drill downs provide an ability to learn more about presented custom metric. This can be helpful if you need to get detailed information about the calculation at a selected point on the chart.

  1. Enter PerfQL query to form your Drill Down. You can enter PerfQL query and create a Drill Down for each series that exists on the chart. To do this, switch the tab in the PerfQL area, enter your query and save.

PerfQL query must contain a condition to display the drill-down when clicking on the chart.

To do this, add a condition to your PerfQL query so that the entity on the x-axis is equal to the clicked_x_value variable, for example:

select key, summary, url from Tickets where resolved = clicked_x_value

Attention, clicked_x_value variable is used only for the drill-down display condition and cannot be used for sorting, filtering, and other actions.

Pay attention to the section Interim metric results in the left menu. In this section you can see the Common Table Expressions (CTE) created for the current custom metric and its attributes that can be included in the drill down.

It is recommended to use the WITH clause and create a PerfQL Query of custom metric using CTE instead of using subqueries. This will help later create more detailed drill-down and display the necessary data.

For adding a clickable link use the following format: [ link title ] ( link URL ). Example: select concat('[',key,'](',url,')') as URL

  1. Click Run Preview to see a preview of the Drill Down

By default, you will see drill down preview at the first point of the Х- axis of the chart. You can change the value of the X-axis, to do this, click on the values of the Axis (X) value field and select a different value.

This step is optional and could remain unfilled

Publish metric

By default, metrics are saved in draft state. To add a metric to the dashboard and aggregate it, the user should publish the metric. This action confirms that the metric is completed and ready.

To publish the metric, click Publish in step 4.

Edit Custom Metric

On the Delivery Metrics page, for the required project:

  1. Click Add → Metrics on the Delivery Metrics. The Metric Catalog is opened. 

 2. Find Custom Metric category in the left navigation menu and open Custom Metric category

  1. Click the gear icon for the necessary metric and click Edit

If a user clicks 'Edit,' it locks the metric for editing by other users until the changes are either published or deleted by the editor. The name of the editor is displayed in the notification message about the metric lock.
The metric will be automatically unlocked after two business days.

  • Until the user publishes updates, the metric widget will display the previously published version.

  • When a user starts editing the metric, a draft is created (a metric with the same name and a ‘Draft' label appears in the Metric Catalog). It is visible only to that user. The original metric is marked with the 'Locked’ label until the daft is published. After publication, the changes from the draft are merged with the previously published version, and the metric is updated.

Copy Custom Metric to other units

The copying of metrics to other units is possible via the Support Team (AskDelivery@epam.com) .

However, please note that only the approach (PerfQL and metric settings) is copied; datasources and CSV files are not included.

Previewing metric’s PerfQL query

Previewing a metric enables all users, not just Key Staff, to understand how the metric is calculated. This is particularly useful for investigating PerfQL queries, which form the basis of metric calculations, without having to edit the metric itself.

  1. Find the custom metric published on the Metric Catalog or a widget on a dashboard that interests you.

  2. Select the "Preview PerfQL" button in the three-dot menu.

  3. A popup will appear with two tabs: "Data Selection" and "Drill Down."

Dear Users,

We want to emphasize that while copying code can be a helpful resource, there is no guarantee that it will work seamlessly for every project. Projects can vary greatly in their specific requirements, dataset structures, and column naming conventions. As a result, it's possible that you may need to adapt the code to suit your project's unique needs. This adaptability is an essential part of ensuring that your project is tailored to its specific context. If you have any questions or need assistance with code adaptation, feel free to reach out to our support team.

Dublicate Custom Metric

If there is a need to create a similar metric with modifications or to experiment with data without compromising the integrity of the original metrics, a duplication of the custom metric can be made. "Save a Copy" will save a duplicate metric for you to modify and publish without the need to manually copy metric settings step by step.

  1. Navigate to the Metric Catalog.

  2. Find the published custom metric widget you want to replicate and click the three-dot menu linked to the selected metric.

  3. In the dropdown menu, select the "Save a copy" option.

  4. A copy of the metric in the draft state is created.

The duplicated metric will be named automatically, following a structured convention that appends the original metric's name with the current date and time in the format "yyyymmdd_hhmmss_utc." It is recommended to rename the metric during the first step of the metric Edit mode.

Delete Custom Metric

To delete Custom Metric from the Metric Catalog: 

  1. Open required project on the Delivery Metrics page.

  2. Click Add → Metrics on the Delivery Metrics. The Metric Catalog is opened. 

  3. Find Custom Metric category in the left navigation menu and open Custom Metric category

  4. Click the gear icon for the necessary metric and click Delete

Copyright © 2024 EPAM Systems, Inc.

All Rights Reserved. All information contained herein is, and remains the property of EPAM Systems, Inc. and/or its suppliers and is protected by international intellectual property law. Dissemination of this information or reproduction of this material is strictly forbidden, unless prior written permission is obtained from EPAM Systems, Inc.

  • No labels