search

shevandrin / rqti

Create QTI Exercises and Exams from R
https://shevandrin.github.io/rqti/
GNU General Public License v3.0
4 stars 2 forks source link

extract_result for 2024 exam #197

Open johannes-titz opened 3 months ago

johannes-titz commented 3 months ago

Calling extract_results("data-raw/mlehreI2024TEST.zip", "items") on 2024 exam gives:

Error in data.frame(file = rep(file_name, length(identifier)), date = rep(test_dt,  : 
  arguments imply differing number of rows: 48, 40

Works for other exams.

I think you have access to the 2024 exam, right?

shevandrin commented 3 months ago

I have checked the exams platform, I looked at Methodenlehre I, but I could not find where i can download zip

johannes-titz commented 3 months ago

On tucexam? You must go to "Bewertungen" and then on the right-hand-side there should be some export; alternatively go to archive. It is a bit hidden.

shevandrin commented 3 months ago

I think it is due to "lagemaßeDavisS2024", german letter in file name, it gives a problem with unzipping

johannes-titz commented 3 months ago

Oh no; once again, I guess I forgot about that. Can we give a meaningful error?

shevandrin commented 3 months ago

there is another issue, i am working on it

shevandrin commented 3 months ago

The problem was with results for Ordering type, when candidate did not provide the answer at all. I have fixed it. Can you check again and give me your error message now?

johannes-titz commented 3 months ago

Great! Works for me. Is this an additional problem to ß? Or was ß not a problem in the first place?

shevandrin commented 3 months ago

if you do not have an error, it means that problems with german letters are only on my side (i think due to win)

shevandrin commented 3 months ago

Great! Works for me. Is this an additional problem to ß? Or was ß not a problem in the first place?

i had both errors: 1) with incorrect parsing Ordering outcomes 2) with german letter in file name. The first is solved, the second i still have

johannes-titz commented 3 months ago

I have no errors. Ok, I see. Would it be an option to replace these values during processing? I am not so much concerned for our data since we stopped using umlauts, but for old data or other people's data it could still be a problem. If it is not possible, we could just give a more meaningful error message. It is not high priority though.

shevandrin commented 3 months ago

I added checker that stops processing and gives a file name. Can you launch it again? I am curious would it give you an error or not? If it stops on your side, I change checker to warning message

johannes-titz commented 3 months ago

I get no error.

shevandrin commented 3 months ago

image in cases like mine a user gets this message we can close this

johannes-titz commented 3 months ago

We could leave it open; if we have time we can try to rename the files automatically.