1.0.78

UDP Marts

Taskforce, Tool Usage Metrics, and Student Activity Score Logic/Calc Changes

Version

1.0.78

Key Features

Taskforce Marts

Adding considerations for academic_session dates when resolving week_in_term numbers. Making this logic consistent with the student activity score and other UDP data marts.
Introducing a new helper table called mart_helper.context__taskforce__assignments_submissions that helps pre-aggregate assignment and submission data. Within this helper table, we introduce an is_quiz boolean field to designate whether assignments and submissions are quizzes.
Introducing _cumulative_ fields for submission scoring fields (e.g. avg_published_score now has a cumulative variant called avg_published_score_cumulative). The non-cumulative field only has scope of its week_in_term. The corresponding cumulative fields have scope of the its week_in_term and all previous weeks.

Tool Usage Metrics

Introduced new helper table called mart_helper.event__general__tool_usage_preagg that helps preaggregate the Caliper tool usage data. This allows us to make some schema and granularity improvements to the final mart.
We introduce 1-hour, 6-hour, and 12-hour metric fields in mart_general.tool_usage_metrics. This now allows usage tracking of each Caliper-emitting tool to be done at an hourly level. The datamart now updates once-per-hour instead of once-per-day.
A new flag called low_hourly_events_flag that signals if we see strangely low activity in the last hour for a tool. The logic for setting the values of this flag is tiered and explained in the mart documentation page linked above. This flag can be leveraged for any reporting or alerting; however, it is a downstream calculation and is not on its own authoritative to denote an outage.

Student Activity Score

We removed a filter to require sis_ext_id for students to be included in the student activity score marts. Removing this filter makes these marts more useful for tenants without SIS data ingested into the UDP.
We removed a hard filter for requiring 10 enrolled students in courses/course_sections to compute a student activity score. We now added a boolean flag called course_has_10_or_more_students in the course_final and course_section_final table. If users wish to replicate results that stay true to the 10-student requirement, including WHERE course_has_10_or_more_students = 1 in user queries will make that happen.
The 10-enrolled student filter is still enforced in the final mart table. The scope of this mart is beyond any single course, so we decided to keep the 10-student filter for the final mart table.

PreviousUDP Marts Release Notes Next1.0.77

Last updated 11 hours ago

Version

1.0.78

Key Features

Adding considerations for academic_session dates when resolving week_in_term numbers. Making this logic consistent with the student activity score and other UDP data marts.

Introducing a new helper table called mart_helper.context__taskforce__assignments_submissions that helps pre-aggregate assignment and submission data. Within this helper table, we introduce an is_quiz boolean field to designate whether assignments and submissions are quizzes.

Introducing _cumulative_ fields for submission scoring fields (e.g. avg_published_score now has a cumulative variant called avg_published_score_cumulative). The non-cumulative field only has scope of its week_in_term. The corresponding cumulative fields have scope of the its week_in_term and all previous weeks.

Introduced new helper table called mart_helper.event__general__tool_usage_preagg that helps preaggregate the Caliper tool usage data. This allows us to make some schema and granularity improvements to the final mart.

We introduce 1-hour, 6-hour, and 12-hour metric fields in mart_general.tool_usage_metrics. This now allows usage tracking of each Caliper-emitting tool to be done at an hourly level. The datamart now updates once-per-hour instead of once-per-day.

A new flag called low_hourly_events_flag that signals if we see strangely low activity in the last hour for a tool. The logic for setting the values of this flag is tiered and explained in the mart documentation page linked above. This flag can be leveraged for any reporting or alerting; however, it is a downstream calculation and is not on its own authoritative to denote an outage.

We removed a filter to require sis_ext_id for students to be included in the student activity score marts. Removing this filter makes these marts more useful for tenants without SIS data ingested into the UDP.

We removed a hard filter for requiring 10 enrolled students in courses/course_sections to compute a student activity score. We now added a boolean flag called course_has_10_or_more_students in the course_final and course_section_final table. If users wish to replicate results that stay true to the 10-student requirement, including WHERE course_has_10_or_more_students = 1 in user queries will make that happen.

The 10-enrolled student filter is still enforced in the final mart table. The scope of this mart is beyond any single course, so we decided to keep the 10-student filter for the final mart table.