from scipy import optimize
def fitting_f(xvalues, *params):
yvalues = []
for x in xvalues:
y = 1/(1+ 2.718281**(params[1]*x+params[0]))
yvalues.append(y)
return numpy.array(yvalues)
X,Y = [], []
for cellsource in xarr:
X.extend(xarr[cellsource])
Y.extend(yarr[cellsource])
print cellsource, 'n=', len(xarr[cellsource])
print 'spearman', stats.spearmanr(X, Y)
coefficients_p, pcov = optimize.curve_fit(fitting_f, X, Y, [0,-1])
print 'line fit coefficients', coefficients_p
x_line = numpy.arange(math.floor(min(X)), math.ceil(max(X)), 0.1)
pylab.plot(x_line, fitting_f(x_line, *coefficients_p), 'k-')
if o.lineparams is not None:
pylab.plot(x_line, fitting_f(x_line, *o.lineparams), 'g-')
# main plotting
import random; random.seed(12)
for cellsource in xarr:
pylab.plot(xarr[cellsource], yarr[cellsource], 'o', label=cellsource, fillstyle='none', markeredgecolor=dr_tools.randomcolour(), markersize=5, markeredgewidth=2)
pylab.legend(loc='lower right')
pylab.ylabel('biallelic fraction')
pylab.xlabel('polyA+ RNA molecules, log2')
pylab.savefig(o.pdfout)
y = 1 / (1 + 2.718281**(params[1] * x + params[0]))
yvalues.append(y)
return numpy.array(yvalues)
X, Y = [], []
for cellsource in xarr:
X.extend(xarr[cellsource])
Y.extend(yarr[cellsource])
print cellsource, 'n=', len(xarr[cellsource])
print 'spearman', stats.spearmanr(X, Y)
coefficients_p, pcov = optimize.curve_fit(fitting_f, X, Y, [0, -1])
print 'line fit coefficients', coefficients_p
x_line = numpy.arange(1, 9, 0.1)
pylab.plot(x_line, fitting_f(x_line, *coefficients_p), 'k-')
import random
random.seed(100)
for cellsource in xarr:
pylab.plot(xarr[cellsource],
yarr[cellsource],
'o',
label=cellsource,
fillstyle='none',
markeredgecolor=dr_tools.randomcolour(),
markersize=5,
markeredgewidth=2)
pylab.legend(loc='lower right')
pylab.ylabel('biallelic fraction (>=20 rpkm)')
pylab.xlabel('mRNA amount, log2 (1(nonlog)=same as ERCC amount)')
pylab.savefig('ERCCsum_vs_biallelic_nonDia_min20rpkm_v6.pdf')
raise Exception
if '__main__' == __name__:
parser = argparse.ArgumentParser()
parser.add_argument('-i', '--csv_in', nargs='*', action='append', required=True)
parser.add_argument('-m', '--marker', '--channel', default='CD57')
parser.add_argument('-o', '--plot_out', default='distribution.pdf')
parser.add_argument('--base_folder_in', default='')
parser.add_argument('--xlim', type=float, default=3)
parser.add_argument('--ylim', type=float)
parser.add_argument('--step', type=float, default=0.01)
parser.add_argument('--noncumulative', action='store_true')
o = parser.parse_args()
random.seed(80)
for files_in, bg_shade in zip(o.csv_in, shades_gen()):
for csv_in in files_in:
df = pandas.read_csv(os.path.join(o.base_folder_in, csv_in))
values = list(df[o.marker])
xarr, yarr = dr_tools.bin(values, start=o.step, end=o.xlim, step=o.step, cumulative=0 if o.noncumulative else -1, fractions=True)
pyplot.plot(xarr, [y*100 for y in yarr], '-', color=dr_tools.mixcolours([bg_shade, dr_tools.randomcolour()], [0.5, 0.5]))
pyplot.xlabel('Signal' if o.noncumulative else 'Cutoff')
pyplot.ylabel('% of cells in bin' if o.noncumulative else '% of cells above cutoff')
if o.ylim is not None: pyplot.ylim(0, o.ylim)
pyplot.title(o.marker)
pyplot.savefig(o.plot_out)
def shades_gen(): for c in ['#ff0000', '#0000ff', '#00ff00']: yield c while True: yield dr_tools.randomcolour()
time_since_baseline_per_sample[patient][visit]
for visit in xarr
]
except KeyError, ValueError:
if not o.quite: print patient
continue
if o.interpolate_points is not None and not o.show_splines:
f_splines = interpolation(xarr, yarr)
xarr, yarr = o.interpolate_points, f_splines(o.interpolate_points)
#for a, b, c in slidingwindow(yarr, 3):
# if (b-a)*(c-b) > 0: direction_continuation.append(1)
# elif (b-a)*(c-b) < 0: direction_continuation.append(0)
if o.colour_patients is None:
c = dr_tools.randomcolour()
elif o.colour_patients[0].endswith('.txt'):
for p in dr_tools.splitlines(o.colour_patients[0]):
if p[0] == patient:
c = {'1': '#3465a4', '0': '#73d216', '2': '#d9b98e'}[p[1]]
break
else:
c = '#ff8888'
#c = dr_tools.mixcolours([c, dr_tools.randomcolour()], [0.8, 0.2])
else:
c = '#5555ff' if patient in o.colour_patients else '#ccffcc'
c = dr_tools.mixcolours([c, dr_tools.randomcolour()], [0.67, 0.33])
if bins:
for b in bins:
b.add1within(xarr, yarr, patient)
else: