Changing Student Anon ID Dump to return Course Specific IDs.

[stephensanchez]

Tests verified, manually tested.

@cpennington Regarding our conversation this morning around the CSV action to pull anonymous user IDs, this PR would change the CSV to include the course-specific IDs.

@Lyla-Fischer This change alters the contract of what is returned by this instructor dashboard action, is there any risk of impacting staff / instructors who may rely on the old Anonymous IDs?

@wedaly

[wedaly]

I'm concerned about this change. Researchers might be relying on the anonymous IDs in these spreadsheets to be consistent over time. I'm also pretty sure there are XModules that are still using IDs that are not unique to a particular course.

I think adding an extra column would be safer.

[Lyla-Fischer]

++ @wedaly

[stephensanchez]

I could definitely add another column for Course Specific Anonymized User IDs. It means I'll likely have to make separate queries.

@cpennington does that sound good to you?

[cpennington]

You can actually just put the return value in these definitions:

@patch.object(instructor.views.api, 'anonymous_id_for_user', Mock(return_value=41))

[stephensanchez]

• Removed Course ID column (redundant)
• Updated tests
• Added Course Specific Anonymized User ID column, instead of changing current column.

[stroilova]

Hello, may I ask @brianhw / @rocha / @mulby whether this new format is parallel to the Anonymous ID info we are currently exporting as part of the research data packages?

• do we export both the global and course-specific anonymous user ids ? user-id-map seems to contain only one of the two, but I remember we had to patch in the other before?
• should the format of the file we provide (user-id-map ) match the format of the file available from the Instructor Dashboard to preempt future questions?

Thank you

[mulby]

@stroilova - we are not currently exporting this ID. See https://edx-wiki.atlassian.net/browse/AN-320.

We export database tables in a very specific format, I doubt we would want to change that for this reason. Also we typically try to mirror the table schema as closely as possible, so we typically aren't joining in columns from other tables etc, which would be necessary to add this information to the current file.

[wedaly]

👍

[stroilova]

Ah ok. That makes sense, thank you @mulby .

On Wed, May 14, 2014 at 3:04 PM, Stephen Sanchez
notifications@github.comwrote:

Merged #3703 https://github.com/edx/edx-platform/pull/3703.

—
Reply to this email directly or view it on GitHubhttps://github.com/edx/edx-platform/pull/3703#event-121162966
.