• Home
  • About
    • Anirban photo
    • About Me
    • Email
  • Blog Posts
    • Writings
    • Tags
  • Skill Set

data.table work diary: Sep 30 - Oct 4

  • Sep 30 - Oct 2, ‘24:

1) Tried to revise the forder caching atime test and observe anticipated results. Learned that the performance improvement was specific to forderv and with datatable.forder.auto.index option set to TRUE. After some experimenting both locally and on my fork’s PR, I managed to create a test case that shows improvement from caching and reusing an index. (16 hours)

2) Reviewed PRs in data.table (such as #6551 and #6553), and one (#50) for the Raft. (4 hours)

  • Oct 3, ‘24:

1) Created a PR (#6555) to introduce the atime test to data.table and made the improvements Toby suggested. (3 hours)

2) Pushed data.table.threads to CRAN again after working on small changes. (2 hours)

3) Updated the blog post on my GHA with the latest changes to my action. (5 hours)

  • Oct 4, ‘24:

1) Going over the weekend, I tried to reproduce the issue Toby is facing with the updated code for the forderv test case - missing historical references. I tried a few changes specific to the introduced code to resolve the issue (I’m certain that it was isolated to this new test case, and likely the test list argument?), but did not come to a solution that fixes the test. (8 hours)