This article provides guidance to troubleshoot issues with Insights reports and metrics.
This article provides guidance to troubleshoot issues with Insights reports and metrics. For issues accessing Insights dashboards, please see the Insights Troubleshooting: Dashboard Access article.
"Report not computable due to improper metric definition" error
While editing a classic report, you may see a “Report not computable due to improper metric definition” error.
Cause
Typically this error occurs for the following reasons:
- A metric is incompatible an attribute (metric-attribute conflict)
- An attribute is incompatible with another attribute (attribute-attribute conflict)
- A metric in the report is set up incorrectly
Metrics and attributes may be incompatible when there is no logical path between two objects in our data model. For instance, the Hours Allocated metric and Task: Name attribute are incompatible because allocated hours are recorded to a project, not a task. Metrics and attributes may also be incompatible if a metric is improperly written.
Resolution
You can identify which metrics and attributes are causing the issue by a process of elimination.
First, remove all the metrics from the report:
- Select the What button to open the list of metrics.
- Uncheck all the metrics.
- Click Done to regenerate the report.
If the error goes away, this means that one of the metrics is conflicting with one or more attributes. Continue to the Resolving a metric-attribute conflict section below.
If the error does not go away, this means that there is a conflict among the attributes. Continue to the Resolving an attribute-attribute conflict section below.
Resolving a metric-attribute conflict
After removing all the metrics from the report, add back them to the report one by one:
- Select the What button to open the list of metrics.
- Check one of the metrics.
- Click Done to regenerate the report.
- If the report regenerates successfully, the metric you added is not the problematic metric.
- Repeat these steps until adding one of the metrics makes the error appear.
Once you identify the metric causing the conflict, remove every metric except that metric from the report. Then remove all the attributes from the report:
- Select the How button to open the list of attributes.
- Uncheck all of the attributes.
- Click Done to regenerate the report.
Now add the attributes back one by one:
- Select the How button to open the list of attributes.
- Check one of the attributes.
- Click Done to regenerate the report.
- If the report regenerates successfully, the attribute you added does not conflict with the metric.
- Repeat these steps until adding one of the attributes makes the error appear.
Once you identify which metric(s) and attribute(s) are conflicting with one another,you can take steps to resolve the issue.
Possible solutions:
- If the metric and attribute are incompatible and cannot be combined in a report, you can break the report up into multiple reports and separate the conflicting metric(s) and attribute(s).
- If the metric and attribute can be combined in a report, the metric needs to be lifted to the correct level. Lifting may require using
SELECT (SELECT SUM())
,BY
, orBY ALL OTHER
statements in the MAQL definition of the metric. For assistance, please reach out to your Business Intelligence Consultant.
Resolving an attribute-attribute conflict
After removing all the metrics from the report, remove all the attributes:
- Select the How button to open the list of attributes.
- Uncheck all the attributes.
- Click Done to regenerate the report.
Add the attributes back back one by one:
- Select the How button to open the list of attributes.
- Check one of the attributes.
- Click Done to regenerate the report.
- If the report regenerates successfully, the attribute you added does not conflict with the other attributes.
- Repeat these steps until adding one of the attributes makes the error appear.
Once you identify which attributes are conflicting with one another, you can break the report up into multiple reports and separate the conflicting attributes.
"Something went wrong" error
While editing a classic report, you may see a "Something went wrong." error.
Resolution
If the report can still be generated, you can typically ignore this error. Click Dismiss to get rid of the error.
Incorrect Metric Values
General Troubleshooting Steps (Classic Reports)
How to See Which Metrics are in a Report and Metric Information
To see which metrics are being used in a classic Insights report, edit the report, then select the What button. The metrics that are being used in the report are listed in the center column that appears.
You can select a metric to see more information about the metric in the right column, such as its description and MAQL definition.
How to See Original Metric Names
Metrics may have been renamed in a report. To see the original names of metrics in a classic Insights report, edit the report, then select Show Configuration. A pane with report information opens. Next, select More Report Info.
In the Metrics Used section is a list of metrics. If a metric was renamed, the original name appears in parentheses.
How to See Report Filters
To see the filters that have been added to a classic Insights report, edit the report, then select the Filters button. The report filters appear in a list below the Filters button.
General Troubleshooting Steps (Dynamic Reports)
How to See Which Metrics are in a Report and Metric Information
To see which metrics are being used in a dynamic Insights report, edit the report, then check the Metrics section(s) to the right of the data catalog.
You can see more information about the metric in the right column—such as its description and MAQL definition—by expanding the metric then hovering over the Info
icon.How to See Original Metric Names
Metrics may have been renamed in a report. To see the original names of metrics in a dynamic Insights report, edit the report, then check the Metrics section(s) to the right of the data catalog. The original name will appear in gray text below the updated name.
How to See Report Filters
To see the filters that have been added to a dynamic Insights report, edit the report, then check the Filters bar which appears below the report name.
Dashboard Filters
Cause
On classic Insights dashboards, dashboard filters may affect the metrics in a report. Dashboard filters override report filters and may cause data to be included in the report that was expected to be excluded by a report filter.
Troubleshooting
To check if dashboard filters are affecting a metric, remove the dashboard filters one by one. Also ensure that there aren't any dashboard filters hidden behind other objects on the dashboard. If a dashboard filter you remove causes a metric to become correct, then that dashboard filter was affecting the metric.
Resolution
Depending on your use case, you may resolve this issue by doing one of the following:
- Editing the report to ignore the conflicting dashboard filter
- Removing the dashboard filter
- Removing the report filter
Example
There is a report that shows billable tracked time by week per person. In timesheets, for the week of Nov 27 - Dec 03, 2022 Angela Jacobs tracked 32 billable hours. However, the report on the Insights dashboard shows that Angela tracked 40 billable hours for that week.
Both the report and dashboard have a Time Entry: Billable filter. The report filter excludes non-billable time from the report. However, because the dashboard filter defaults to All and overrides the report filter, both billable and non-billable time entries are counted, resulting in the 40 hours value for Angela.
After removing the Time Entry: Billable dashboard filter, the report shows Angela tracked 32 billable hours, as expected. The dashboard filter no longer overrides the report filter.
Report Filters
Cause
On a classic or dynamic dashboard, report filters limit the data in a report and affect all the metrics and attributes in the report. An incorrectly added report filter may cause incorrect metric values.
Troubleshooting
To check if report filters are affecting a metric, remove the report filters one by one. If the report filter you remove causes a metric to become correct, then the report filter was affecting that metric.
Resolution
Depending on your use case, you may resolve this issue by removing the report filter or the metric.
Example
There is a report on a dashboard showing tracked time by week per user. In timesheets, Grace Murray tracked 40 hours for the week of Nov 27 - Dec 03, 2022 (including a task that is now archived). However, the report shows that Grace tracked 32 hours for that week.
In the report editor, there is a filter for the Tasks: Archived attribute—this filter excludes data for archived tasks.
After removing the Tasks: Archived report filter, the report on the dashboard now shows Grace tracked 40 hours, as expected. The report filter no longer excludes data.
Other Metrics and Attributes
Cause
On a classic or dynamic dashboard, other metrics and attributes may affect metric values if the metric is defined incorrectly. If adding or removing an attribute to a report changes the metric sum, this typically indicates that the metric does not contain the proper BY
statements (known as “lifting”).
Troubleshooting
To identify which metric or attribute is affecting a metric, add and remove the metrics and attributes one by one and see which one affects the metric.
Resolution
Depending on which metric or attribute is affecting the metric, you may need to add BY
statements to the metric definition. For assistance, please reach out to your Business Intelligence Consultant.
Example
There is a custom metric for invoice amount.
In a report, the Invoice Amount sum changes when the Invoice: Number attribute is added to the report.
This indicates that the custom metric needs to be lifted. The MAQL definition of the Invoice Amount metric is updated to add a BY
statement, which specifies how the metric should be aggregated.
Now, adding Invoice: Number to the report does not affect the Invoice Amount sum.
Metrics within the Metric
Cause
On a classic or dynamic dashboard, a metric may display an incorrect value if the other metrics it is made up of are incorrectly defined.
Troubleshooting
To identify which sub-metric in a metric is incorrect, add the sub-metrics to the report and check each one.
Resolution
Once you identify which sub-metric is incorrect, you can edit the sub-metric and update its MAQL definition.
Example
There is a project custom field called Discount % that takes values from 0 to 100 (as opposed to 0 to 1).
In Insights, there is a custom metric called Discount that multiplies the standard Project Budget metric by a custom metric called Discount %.
In a report, the value of Discount is incorrect—it is much larger than the project budget.
To pinpoint which part of the Discount metric is incorrect, we add the Discount % sub-metric to the report. The values of Discount % are not between 0 and 1, as required for multiplication, but are between 0 and 100 instead. The Discount % sub-metric is incorrectly defined.
The Discount % custom metric is updated to divide the Discount % custom field by 100, to produce values between 0 and 1 as required for multiplication.
The Discount % and Discount metrics in the report are now correct.
Metric Filters
Cause
On a classic or dynamic dashboard, a metric may display an incorrect value due to incorrectly defined metric filters in the MAQL definition. Metric filters may cause data to be included or excluded unexpectedly.
Troubleshooting
To check if metric filters are affecting a metric, open the metric details and look for the following types of MAQL operators and expressions:
- Operators:
=
,!=
,<>
,>
,<
,>=
,<=
-
BETWEEN
orNOT BETWEEN
-
IN
andNOT IN
WITH PF EXCEPT
-
THIS
,PREVIOUS
,NEXT
-
LIKE
,NOT LIKE
,ILIKE
,NOT ILIKE
Resolution
You may resolve this issue by updating the MAQL definition of the metric. For assistance, please reach out to your Business Intelligence Consultant.
Example
In Task Tracker, the actual fees of started tasks in a certain project are $720. In an Insights report, there is a custom metric that displays actual fees of started tasks. The actual fees of started tasks for that project are shown as $120 instead of $720.
In the custom metric’s definition, we see that the MAQL is incorrect—the metric filters for completed tasks instead of started tasks.
The metric’s MAQL definition is updated. The metric now filters for started tasks.
The metric now displays the correct actual fees of started tasks.
Empty Metric Values
Cause
On a classic or dynamic dashboard, a metric may show no values if any part of its calculation returns a null value, regardless of whether the sub-metrics in it have values.
Troubleshooting
To check if a metric may be showing no values due to null values in its calculation, check if metric has sub-metrics and no IFNULL
statements in its MAQL definition.
Resolution
To resolve this issue, update the metric's MAQL definition and add IFNULL
statements around the sub-metrics to handle the null values. For assistance, please reach out to your Business Intelligence Consultant.
Example
There is a custom metric that displays either allocated hours or actual hours.
In a report, the custom metric displays no value for some projects, despite those projects having values for actual hours or allocated hours.
The custom metric is updated to add IFNULL
statements.
The custom metric now populates values for all projects.
Comments
2 comments
This article has been updated to include a video.
This article has been updated to include an additional video in the General Troubleshooting Steps (Classic Reports) section.
Please sign in to leave a comment.