def main():
option_parser, opts, args =\
parse_command_line_parameters(**script_info)
map_data, map_header, map_comments = parse_mapping_file(open(
opts.map, 'U'))
map_dict = mapping_file_to_dict(map_data, map_header)
distdict = parse_distmat_to_dict(open(opts.distance_matrix, 'U'))
if opts.colorby == None:
colorby_cats = [None]
else:
colorby_idx = map_header.index(opts.colorby)
colorby_cats = list(set([map_data[i][colorby_idx] for\
i in range(len(map_data))]))
textfilename = os.path.splitext(opts.output_path)[0] + '.txt'
text_fh = open(textfilename, 'w')
text_fh.write(opts.axis_category + '\tdistance\tSampleID' + '\n')
colorby_cats.sort()
plt.figure()
for cat_num, cat in enumerate(colorby_cats):
# collect the primary and secondary samples within this category
state1_samids, state2_samids = get_sam_ids(map_data, map_header,
opts.colorby, cat,
opts.primary_state,
opts.secondary_state)
state1_samids =\
list(set(state1_samids).intersection(set(distdict.keys())))
state2_samids =\
list(set(state2_samids).intersection(set(distdict.keys())))
if state1_samids == [] or state2_samids == [] or \
(len(state1_samids) == 1 and state1_samids == state2_samids):
raise RuntimeError("one category of samples didn't have any valid"+\
" distances. try eliminating samples from -p or -s, or changing"+\
" your mapping file with filter_samples_from_otu_table.py")
# go through dmtx
state1_avg_dists = get_avg_dists(state1_samids, state2_samids,
distdict)
# plot
xvals = [float(map_dict[sam][opts.axis_category]) for\
sam in state1_samids]
try:
color = plt.cm.jet(cat_num / (len(colorby_cats) - 1))
except ZeroDivisionError: # only one cat
color = 'b'
plt.scatter(xvals,
state1_avg_dists,
edgecolors=color,
alpha=.5,
facecolors='none')
plt.xlabel(opts.axis_category)
plt.ylabel('average distance')
lines = [str(xvals[i])+'\t'+str(state1_avg_dists[i])+\
'\t'+state1_samids[i]+'\n' for i in range(len(xvals))]
text_fh.writelines(lines)
if opts.colorby != None: plt.legend(colorby_cats)
plt.savefig(opts.output_path)
def test_get_avg_dists(self):
"""get_avg_dists functions as expected """
dmtx_str = StringIO.StringIO("""\ts1\ts2\ts3
s1\t0\t.5\t.6
s2\t.5\t0\t.7
s3\t.6\t.7\t0.0
""")
distdict1 = parse_distmat_to_dict(dmtx_str)
distdict2 = {
's1': {
's2': .5,
's3': .6
},
's2': {
's1': .5,
's3': .7
},
's3': {
's2': .7,
's1': .6
}
}
state1_samids = ['s1', 's2']
state2_samids = ['s3', 's2']
# note s2 in both
exp_avgs = [.55, .7]
obs_avgs = get_avg_dists(state1_samids, state2_samids, distdict1)
assert_almost_equal(exp_avgs, obs_avgs)
def main():
option_parser, opts, args =\
parse_command_line_parameters(**script_info)
map_data, map_header, map_comments = parse_mapping_file(
open(opts.map, 'U'))
map_dict = mapping_file_to_dict(map_data, map_header)
distdict = parse_distmat_to_dict(open(opts.distance_matrix, 'U'))
if opts.colorby is None:
colorby_cats = [None]
else:
colorby_idx = map_header.index(opts.colorby)
colorby_cats = list(set([map_data[i][colorby_idx] for
i in range(len(map_data))]))
textfilename = os.path.splitext(opts.output_path)[0] + '.txt'
text_fh = open(textfilename, 'w')
text_fh.write(opts.axis_category + '\tdistance\tSampleID' + '\n')
colorby_cats.sort()
plt.figure()
for cat_num, cat in enumerate(colorby_cats):
# collect the primary and secondary samples within this category
state1_samids, state2_samids = get_sam_ids(map_data, map_header,
opts.colorby, cat, opts.primary_state, opts.secondary_state)
state1_samids =\
list(set(state1_samids).intersection(set(distdict.keys())))
state2_samids =\
list(set(state2_samids).intersection(set(distdict.keys())))
if state1_samids == [] or state2_samids == [] or \
(len(state1_samids) == 1 and state1_samids == state2_samids):
raise RuntimeError("one category of samples didn't have any valid" +
" distances. try eliminating samples from -p or -s, or changing" +
" your mapping file with filter_samples_from_otu_table.py")
# go through dmtx
state1_avg_dists = get_avg_dists(
state1_samids,
state2_samids,
distdict)
# plot
xvals = [float(map_dict[sam][opts.axis_category]) for
sam in state1_samids]
try:
color = plt.cm.jet(cat_num / (len(colorby_cats) - 1))
except ZeroDivisionError: # only one cat
color = 'b'
plt.scatter(xvals, state1_avg_dists, edgecolors=color, alpha=.5,
facecolors='none')
plt.xlabel(opts.axis_category)
plt.ylabel('average distance')
lines = [str(xvals[i]) + '\t' + str(state1_avg_dists[i]) +
'\t' + state1_samids[i] + '\n' for i in range(len(xvals))]
text_fh.writelines(lines)
if opts.colorby is not None:
plt.legend(colorby_cats)
plt.savefig(opts.output_path)
def test_get_avg_dists(self):
"""get_avg_dists functions as expected """
dmtx_str = StringIO.StringIO("""\ts1\ts2\ts3
s1\t0\t.5\t.6
s2\t.5\t0\t.7
s3\t.6\t.7\t0.0
""")
distdict1 = parse_distmat_to_dict(dmtx_str)
distdict2 = {'s1':{'s2':.5,'s3':.6},'s2':{'s1':.5,'s3':.7},
's3':{'s2':.7,'s1':.6}}
state1_samids = ['s1','s2']
state2_samids = ['s3','s2']
# note s2 in both
exp_avgs = [.55, .7]
obs_avgs = get_avg_dists(state1_samids,state2_samids,distdict1)
self.assertFloatEqual(exp_avgs, obs_avgs)