Resolved Issues (5.0.0.0)
These are the resolved issues for the Performance Analytics and Intelligence 5.0.0.0 release.
JIRA Number
Case Number
Summary
PAI-8241
16926817
The error metrics in ML time series forecasting are now computed from first available slice date of Servigistics archived forecast within the validation period decided by train-test split. The training data is adjusted accordingly.
PAI-8264
16936314
Filters for part, location and forecast stream on SKU Forecast Analysis page are working correctly and data on report is populated based on filtered criteria.
PAI-8267
N/A
The performance of the Planner Workload dashboard was improved.
PAI-8270
16935827
The New Part Introduction Forecast dashboard no longer has a dependency on SFTP containers. All of the widgets can now display results even if the SFTP container is turned off.
PAI-8351
16976943
The Review Reason name is now displayed on the Top Review Reasons and Review Reason Summary widgets on the Performance 360 dashboard.
PAI-8402
16947516
17025902
The Demand Forecast Override Analysis dashboard is now generated correctly on a Snowflake platform.
PA-8780
N/A
The Manual System Date was incorporated into Machine Learning Time Series Forecasting, to ensure that the process aligns with the system date used by the Best Fit process in Servigistics.
PAI-8808
17111479
The performance of the Analytics Autopilot processes for Snowflake-based PAI can be improved when archiving tables containing more than 1GB of data by loading the data into Snowflake in multiple files rather than in one file. A new system property named servigistics.analytics.sf.csvFileSplitRowCount can be added to the AutopilotServer.properties file to specify the maximum number of rows to include in each file. The recommended setting is 500000.
PAI-8813
17121772
Tables that are configured to be archived in PAI can now contain numeric columns that don't have a scale specified.
PAI-8819
16976864
The Forecast Error Analysis and Forecast Error Analysis - Advanced dashboards are now loaded properly.
PAI-9073
17162032
The data in the Forecast Accuracy By Demand Category and Forecast Method Count of the Forecast Analysis dashboard no longer disappears after being displayed for approximately 15 minutes. The data is displayed as expected.
PAI-9253
17207564
The QO_PAI_PlannerPerformance Query Object that is used to build the Planner Workload dashboard was improved. The associated Analytical Object can now be built successfully when there are more than 135 users.
PAI-9374
17300519
The forecast accuracy calculation was corrected for cases when the average demand is zero.
PAI-9466
17303321
Instructions for updating Intellicus to a new server were added.
PAI-9493
17279349
17316915
The Planning Analytics Generator runs successfully after making a scenario production.
Was this helpful?