[feature] : Consider using compression when saving to .topostats file

[SylviaWhittle]

Is your feature request related to a problem?

Currently we don't use compression when saving to .topostats files, resulting in huge file system weight. My own laptop is 90% full, mostly from .spm and .topostats files.

Describe the solution you would like.

At a cost of I/O speed, we could enable compression of our HDF5 datasets when saving to .topostats files.

Here's some documentation on compression when saving to .HDF5 files using h5py.
https://docs.h5py.org/en/latest/high/dataset.html

I haven't looked into this further but I did do a little investigation a while back and saw that compressed .HDF5 files were impressively small.

Note that array_size here is the size of one side of the square, so array_size = 100 means a 100 x 100 image.

Also note that the load time differences for anything below a 2000 x 2000 pixel image are negligible in my test case.

Describe the alternatives you have considered.

No response

Additional context

No response

[ns-rse]

Hadn't realised there was a compression option to HDF5 formats.

Presumably a similar overhead is incurred when writing to compressed formats.

No reason why not to add this but it should be configurable both in terms of which compression method to use (both gzip of lzf are available according to documentation and the former allows users to specify compression levels) and whether to compress at all. Whether or not that default should be compressed or not is something to consider, always compressing may make .topostats files less portable (as a simple example I'm unsure if h5glance works with compressed HDF5 files).