Notes on Issue #48 - How Resilient Are Males vs Females

[clarrain]

Current Plan:
Look in table female_first_pulls of Sandbox - match the pull_request_id with pull_request_id from table pull_request_history
Save the pull_request_id and created_at of instances where action "closed" exists
Match the pull_request_id from pull_request_history with pull_request_id from pull_requests
Save the base_repo_id or head_repo_id for each of these _pull_request_id_s from pull_requests
Look in pull_request_history for instances where actor_id equals user_id from female_first_pulls after the created_at date saved from pull_request_id
For the above instances, save the pull_request_id
Go to pull_requests and see if this pull_request_id has the same head/base repo id. If yes, count the instances of this match

[clarrain]

Make a new table instead of relying on female_first_pulls - start w/ g_pull_closes in Sandbox

[CaptainEmerson]

You might consider using views or temporary tables. Views can be treated like tables, but they're actually just queries. Temporary tables are like normal tables, but disappear after you disconnect your MySQL session.

[CaptainEmerson]

You'd want to use a temporary table instead of a view when the view would take a long time (say, more than a few minutes) to execute.

[clarrain]

Okay, thanks! I was planning on using views for sure, and I'll look into instances where using a temp table would be better suited

[clarrain]

pull_request_history not accurate - re-write plan using g_pull_closes

[clarrain]

I'm changing up how I want to go about finding this out - instead of looking at the statistics of a pull request creator for a specific project, I'm doing to look at the general statistics of a creator - after their first "closed" pull request, do they make any more requests (how many? What is the state of their final pull request - closed or open?) for each gender. After working with Denae on the revised plan, I realized that it would be nearly impossible to limit the stats to one specific project with joins using only the data that g_pull_closes has.

[CaptainEmerson]

Please make your technical notes on issue #48 itself. This repository is public, the other is private. We certainly can't post any data here, and probably don't want to post metadata (e.g., our tables) either.

[clarrain]

Got it!