search

aaowens / PSID.jl

Quickly assemble data from the Panel Study of Income Dynamics (PSID)
MIT License
25 stars 9 forks source link

Updating files and hashes #52

Open economoser opened 2 weeks ago

economoser commented 2 weeks ago

I noticed that a couple of files and hashes may need updating:

  1. The PSID codebook is now available (again) on the official website at https://simba.isr.umich.edu/downloads/PSIDCodebook.zip, but that file differs from that in your Google Drive, which produces a warning based on hash mismatch, which in the old (i.e., pre-2021) version based on your Google Drive file did not occur. Which file and hash should we use for the 2021 version of the PSID data?
  2. I get a warning notifying me that the hash for file fam2017er.zip is 514a32736f62d40b776f0056b67113641cf704e584895bc9663de6c0e87de54e, but that the program expected 5ade1a3f42ed84c892fe8ff16365b85b0dc84ac66f4d454e291af12008e9b35d (accessed in ~/src/init.jl:19).
  3. For what it's worth, when using the old (i.e., pre-2021) version of your code, I also got a warning notifying me that the hash for file ind2019er.zip is aec6362d407925edc128016ed26d51184721e2cd693a63bf2e280056a4d056a7, but that the program expected 583e541f2f8d8c9ea12e2f53e8f7594f525d95ad929f5ce804d030e1a593991d.

When running the new (i.e., including 2021) version of your code, I also encounter the following error: ERROR: LoadError: ind2019er.zip not found and is required. I thought that the new version for the 2021 wave of the PSID data should only need ind2021er.zip and not the earlier versions but maybe I'm mistaken?

Again, thanks a lot for all your help!

aaowens commented 2 weeks ago

That's surprising. There's no mention of ind2019er.zip anywhere in the code. You sure it's the latest?

I'll check on the codebook, didn't know it was available again.