search

circstat / pycircstat

Toolbox for circular statistics with Python
MIT License
157 stars 42 forks source link

Kuiper Table Installation #63

Open DanielHHowell opened 5 years ago

DanielHHowell commented 5 years ago

Having difficulty with the Kuiper table - receiving this error when trying to run the test: "[Errno 2] No such file or directory: '/home/daniel/anaconda3/lib/python3.7/site-packages/data/kuiper_table.npy'". Tried manually downloading from the repo and placing it in the directory it was looking for but received an axis error so not sure if that was the right thing to do

fabiansinz commented 5 years ago

I guess you made sure that the file was actually placed there? Did you check the file right?

DanielHHowell commented 5 years ago

Yes I tried placing the file there, and I believe the axis error is because N>>500 in the lookup table

dominikstrb commented 3 years ago

I got the same error, because my N was too large. Does anyone know a fix for this?

DanielHHowell commented 3 years ago

I can't remember if I was able to get it working, but looking back at the project it seems for one reason or another I switched to using astropy.stats.kuiper

philippberens commented 3 years ago

In this case, you can just use the last line, and P will be < than the pvalue in the last line.

dominikstrb commented 3 years ago

Thanks. I believe this is what is tried to be achieved here, but it does not work in my case. If I add -1 to that line, it works.

If you would like your library to be pip-installable without the need for users to download additional files, you could use this in your setup: https://python-packaging.readthedocs.io/en/latest/non-code-files.html.

philippberens commented 3 years ago

sorry, i really should get back to being better on taking care of this code... new years resolutions!

A-ep93 commented 2 years ago

Is this (using the value of the last line of the table (N=501) for all larger sizes of arrays) a safe choice for both cases when the hypothesis gets rejected/accepted?