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

data.table work diary: Mar 25 - 29

  • Mar 25, ‘24:

1) Made the changes that Michael suggested here.

2) Created a first draft of a step to check and update for files where OpenMP is used (#31), to be used within the repo-meta-tests.yaml workflow in data.table.

3) Mar 26, ‘24:

4) Collected feedback from Toby, working on discussed things.

  • Mar 27, ‘24:

1) Writing some benchmark code to see which routines in data.table benefit more from parallelization when having more number of rows vs more number of columns for the considered data inputs. Since I can’t test locally on MacOS machine (data.table isn’t multi-threaded for my OS, so no point in setDTthreads(n > 1)), I’m running on an Ubuntu container via a GitHub action. Will share results soon.

2) Attended Toby’s presentation regarding the R project in Google Summer of Code.

3) Completed benchmarks, please see #32. Based on the results I see, I’ll update my PR for the OpenMP documentation accordingly tomorrow.

  • Mar 28, ‘24:

1) Added information to my OpenMP documentation PR after finishing my benchmarking and some research on better speedups for a large number of rows vs columns in the input data.

2) Made some progress towards my blog post on the GitHub Action.

3) Commented on a few issues in data.table.

  • Mar 29, ‘24:

1) Trying to divide the timed segment for the atime step in my action for installation and test runs separately. Problem is, everything inside R -e ' ... ' needs to be run together in one session. So I can’t use the approach I was using. Trying new approaches. Also trying to correct the ‘Before’ label (parent commit of PR #4440’s merge commit).

2) Tested and commented on PR #6034.

3) Made a small PR #6038 based on what Michael suggested and commented on a few other threads.