Generate graphs from worm behavior data.

ViewWormBehaviorData.py

# -*- coding: utf-8 -*-


import sys, re
import h5py
import matplotlib.pyplot as plotter

# This script takes a local filename of a file that was downloaded from the
# C. elegans behavioural database and writes a wormbehavior.html file in
# the current directory with links to .png files that plot the datasets
# from the 'worm' Group. It does nothing with the 'info' Group, which
# contains some info about the type of worm and environment. Note that the 
# .png filenames start with a '.', so you may need ls -a to see them.
# It takes a few minutes to run.
# See https://github.com/openworm/OpenWorm/issues/82 for more info.


def view_group(group1, h5file, html_file, name = ''):
    for fld in group1.keys():
        print("%s/%s %s" % (name, fld, len(group1[fld])))
        name1 = name + '/' + fld
        if isinstance(group1[fld], h5py.Group):
            view_group(group1[fld], h5file, html_file, name = name1)
        else:
            try:
                is_ref = isinstance(group1[fld][0][0], h5py.h5r.Reference)
            except (IndexError, AttributeError):
                is_ref = False
            if is_ref:
                dataset = []
                i = 0
                for ref_set in group1[fld]:
                    i += 1
                    for ref1 in ref_set:
                        # I'm just guessing here about what to do
                        dataset.append(h5file[ref1][0][0])
                        print("deref %2d %.72s" % (i, str(dataset)))
            else:
                dataset = group1[fld]
            print("dataset %s" % dataset[0].__class__)
            print("%.72s" % dataset[:8])
            plotter.close()
            plotter.plot(dataset)
            #plotter.ylabel(?)
            plotter.title(name1)
            fname = re.sub("/", ".", name1)  + '.png'
            plotter.savefig(fname)
            html_file.write('<li><a href="%s">%s</a></li>\n' % (fname, name1))


def view_file(filename):
    try:
        h5file = h5py.File(filename, 'r')
    except IOError:
        print("Problem with file %s" % filename)
        raise
    
    html_file = open('wormbehavior.html', 'w')
    html_file.write("<ol>\n")
    view_group(h5file['worm'], h5file, html_file)
    html_file.write("</ol>\n")

if __name__ == "__main__":
    view_file(sys.argv[1])



# I didn't use pytables because I got this:
#/worm/locomotion/motion/backward/frames/distance (UnImplemented(29, 1)) ''
#  NOTE: <The UnImplemented object represents a PyTables unimplemented
#         dataset present 

A belated thanks for this, Peter!

I've included this in the movement_validation repository now, in case people are interested in the techniques used, such as recursively accessing the HDF5 data and plotting it. See also the issue I raised.