def test_take(self):
"""take returns subset with correct attributes"""
input = dict(counts=[[0, 1], [2, 3], [4, 5], [6, 7], [8, 9]],
info={'abcd': 'efg'},
labels=['0', '1', '2', '3', '4'])
coll = RegionCollection(**input)
sub = coll.take([0, 1, 2])
self.assertEqual(sub.info, input['info'])
self.assertEqual(sub.counts.tolist(), [[0, 1], [2, 3], [4, 5]])
self.assertEqual(sub.labels.tolist(), ['0', '1', '2'])
def main():
option_parser, opts, args =\
parse_command_line_parameters(**script_info)
if ',' not in opts.ylim:
raise RuntimeError('ylim must be comma separated')
ylim = map(float, opts.ylim.strip().split(','))
print 'Loading counts data'
data_collection1 = RegionCollection(filename=opts.collection1)
window_size = data_collection1.info['args']['window_size']
data_collection2 = RegionCollection(filename=opts.collection2)
# filter both
if opts.cutoff < 0 or opts.cutoff > 1:
raise RuntimeError('The cutoff must be between 0 and 1')
data_collection1 = data_collection1.filteredChebyshevUpper(opts.cutoff)
data_collection2 = data_collection2.filteredChebyshevUpper(opts.cutoff)
# make sure each collection consists ot the same genes
shared_labels = set(data_collection1.labels) & \
set(data_collection2.labels)
data_collection1 = data_collection1.filteredByLabel(shared_labels)
data_collection2 = data_collection2.filteredByLabel(shared_labels)
assert set(data_collection1.labels) == set(data_collection2.labels)
if opts.sample_top is None:
sample_top = data_collection1.N
else:
sample_top = opts.sample_top
indices = range(sample_top)
data_collection1 = data_collection1.take(indices)
data_collection2 = data_collection2.take(indices)
print 'Starting to plot'
if opts.bgcolor == 'black':
grid = {'color': 'w'}
bgcolor = '0.1'
vline_color = 'w'
else:
grid = {'color': 'k'}
vline_color = 'k'
bgcolor = '1.0'
vline = dict(x=0,
linewidth=opts.vline_width,
linestyle=opts.vline_style,
color=vline_color)
plot = PlottableSingle(height=opts.fig_height / 2.5,
width=opts.fig_width / 2.5,
bgcolor=bgcolor,
grid=grid,
ylim=ylim,
xlim=(-window_size, window_size),
xtick_space=opts.xgrid_lines,
ytick_space=opts.ygrid_lines,
xtick_interval=opts.xlabel_interval,
ytick_interval=opts.ylabel_interval,
xlabel_fontsize=opts.xfontsize,
ylabel_fontsize=opts.yfontsize,
vline=vline,
ioff=True)
x = numpy.arange(-window_size, window_size)
if opts.metric == 'Mean counts':
stat = averaged
else:
data_collection1 = data_collection1.asfreqs()
data_collection2 = data_collection2.asfreqs()
stat = summed
plot_sample(plot, data_collection1, stat_maker(stat, data_collection1), x,
opts.title, opts.xlabel, opts.ylabel, 'b', opts.legend1,
opts.plot_stderr)
plot_sample(plot, data_collection2, stat_maker(stat, data_collection2), x,
opts.title, opts.xlabel, opts.ylabel, 'r', opts.legend2,
opts.plot_stderr)
plot.legend()
plot.show()
if opts.plot_filename and not opts.test_run:
plot.savefig(opts.plot_filename)
else:
print opts.plot_filename
