- one minute read - 24 words

Just because you can topic model something doesn’t mean it actually tells us anything (and please don’t ever describe computational text analysis as “objective”).

similar posts:

One of those afternoons where I’m auditing someone’s analysis code, but it’s an analysis of 4M rows of data, so I’m also doing spurts of grading while I wait for code to execute.

35 GB of data is a lot to begin with, but when it’s 35 GB of CSVs? That’s when it starts to really register.

I got a reminder today that I do the kind of research where something as hilariously unintuitive as telling a program to treat long numbers as “words made up of 0-9” is actually a critical step to making sure you get the right results.

comments:

You can click on the < button in the top-right of your browser window to read and write comments on this post with Hypothesis.