Open mcfneuro opened 4 years ago
OK cool - tag me when you have a draft and I'll look it over!
OK, I have a first draft and started working of organizing my data (uploaded/updated a short code using a bunch of your suggestions)
Looks great! I like the general format of everything. It might be helpful to mention what kind of subjects you are working with (mice? rats? humans?) and your N. Otherwise, I like it!
One other thing to mention is that, unless it is very small, it's usually not a great idea to store data on your git repo. If you must, it's create to try to compress it as much as possible. But generally you don't want to store large files on github for a number of reasons. Remember also that this is a public repo and anyone can download what you put up.
Looks great! I like the general format of everything. It might be helpful to mention what kind of subjects you are working with (mice? rats? humans?) and your N. Otherwise, I like it!
Done. Thanks
One other thing to mention is that, unless it is very small, it's usually not a great idea to store data on your git repo. If you must, it's create to try to compress it as much as possible. But generally you don't want to store large files on github for a number of reasons. Remember also that this is a public repo and anyone can download what you put up.
Where/how would you store it ? Files are small (2 files of 20 MB each). But, regardless, I'd like to know what would be best practice ?
Where/how would you store it ? Files are small (2 files of 20 MB each). But, regardless, I'd like to know what would be best practice ?
20 MB is actually pretty big for git.. keep in mind that usually you're storing text files and such that rarely get larger than a few MB. Although, once you convert these to numpy arrays, you can use np.savez_compressed() to create very compressed arrays that should be way smaller than those matfiles.
Generally, you store the data locally -- it's the code that you're sharing and versioning on git. If you want to host data online, there are other better ways to do it. If you want to store metadata or map your directories, you can also check out datalad
Just opening an "issue" to document our discussion from yesterday PM