If I remember correctly, read_gudb uses pandas because its load_csv used to be a lot faster than numpy's loadtxt. This changed in numpy 1.23.0 when the function was implemented in C, so I guess we can drop pandas as an optional requirement. I don't think it's necessary to bump the required numpy version to 1.23.0 since it is just a speed improvement, not a functional one, what do you think @cbrnr?
If I remember correctly,
read_gudb
usespandas
because itsload_csv
used to be a lot faster than numpy'sloadtxt
. This changed in numpy 1.23.0 when the function was implemented in C, so I guess we can drop pandas as an optional requirement. I don't think it's necessary to bump the required numpy version to 1.23.0 since it is just a speed improvement, not a functional one, what do you think @cbrnr?