Pass database options (data_path, input_format) as command-line arguments?

[generalciao]

For submitting jobs on a cluster, I would like to be able to execute suite2p from the command-line, passing the experiment-specific parameters as arguments. I think this is mainly "data_path" and "input_format". In some cases there may also be format-specific options (e.g. sbx_ndeadcols). Would it be sensible to add this functionality?

When I currently try to pass these 2 or 3 "database" options using command-line flags, I get an "unrecognized arguments" error. All the other "non-database" options (such as --fs), which are documented in --help, work fine for me.

My current workaround is to execute a one-line python string that creates a db dict and saves it to db.npy, then pass this filepath as the --db argument. echo -e "import numpy\ndb = {'data_path':['test-data'], 'input_format':'sbx', 'sbx_ndeadcols':-1}\nnumpy.save('test-data/db.npy', db)" | python

[carsen-stringer]

yes you need to make a db.npy file to run from the command line directly. if you implement this directly for single folders I'm happy to accept a pull request