def draw_bar(days, START=None):
data = get_freq(days, START)
k = sorted(list(data.keys()))
res = [0] * (k[1])
for i in range(1, len(k) - 1):
res += [data[k[i]]] * (k[i + 1] - k[i])
res += [0] * (1440 - k[-1])
x = np.array(list(range(len(res))))
y = np.array(res)
fig, ax = plt.subplots()
colors = cm.YlGnBu(y / float(max(y)))
ax.bar(x, y, width=1.0, color=colors)
ticks = {}
for hour in range(0, 24, 3):
ticks[hour * 60] = '{}:00'.format(hour)
ticks[24 * 60] = '0:00'
ax.set_xticks(list(ticks.keys()))
ax.set_xticks(list(range(0, 24 * 60, 60)), minor=True)
ax.set_xticklabels(list(ticks.values()))
ax.grid(b=True, which='minor', axis='x', linestyle='dotted')
ax.set_xlim(6 * 60, 24 * 60)
plt.show()
def _create_colors(classes):
n_cts = len(classes)
color_norm = colors.Normalize(vmin=-n_cts / 3, vmax=n_cts)
ct_arr = np.arange(n_cts)
ct_colors = cm.YlGnBu(color_norm(ct_arr))
return ct_colors
def _create_colors(lr):
n_cts = lr.classes_.shape[0]
color_norm = colors.Normalize(vmin=-n_cts / 3, vmax=n_cts)
ct_arr = np.arange(n_cts)
ct_colors = cm.YlGnBu(color_norm(ct_arr))
return ct_colors
def DrawLocalCurrents(self, bDraw=None):
if bDraw==None:
bDraw = self.cbOptions1.lines[2][0].get_visible()
self.axCM.set_visible(bDraw)
if bDraw and self.Gate is not None:
#I = self.Mol.CurrentMatrix(self.Bias, self.Gate)
I = self.Mol.LocalCurrents(self.Bias, self.Gate)
#DisplayMatrix(I)
I = np.real(I)
maxCurrent = abs(I).max()
print "Maximum current:", maxCurrent
for i in range(self.Mol.N):
for j in range(i+1,self.Mol.N):
Inet = (I[i,j]-I[j,i])/2
if abs(Inet)>maxCurrent*0.1:
color = cm.YlGnBu(1-math.log10(abs(Inet))/-10)
if Inet>0:
self.LCPlot[i,j].set_linewidth(Inet/maxCurrent*10)
self.LCPlot[i,j].set_color(color)
self.LCPlot[i,j].set_visible(True)
self.LCPlot[j,i].set_visible(False)
print "Current from", i, "to", j, "is", I[i,j]
else:
self.LCPlot[j,i].set_linewidth(-Inet/maxCurrent*10)
self.LCPlot[j,i].set_color(color)
self.LCPlot[j,i].set_visible(True)
self.LCPlot[i,j].set_visible(False)
print "Current from", j, "to", i, "is", -I[i,j]
else:
self.LCPlot[i,j].set_visible(False)
self.LCPlot[j,i].set_visible(False)
else:
for i in range(self.Mol.N):
for j in range(self.Mol.N):
if i!=j:
self.LCPlot[i,j].set_visible(False)
self.fig.canvas.draw()
# output colormap to txt file
#
# dependencies: matplotlib
from matplotlib import cm
import csv
with open('ylgnbu.txt', 'wb') as f:
writer = csv.writer(f, delimiter=' ')
for c in range(0,255):
# in this case I picked YlGnBu but you can pick any
# color map you want.
writer.writerow(cm.YlGnBu(c))
f.close()
def main(args, tick_font_size, label_font_size):
lines = open(args.input_file).readlines()
data_lines = lines[1:]
experiment_names = lines[0].split()
experiment_index = args.experiment_columns
experiments = [experiment_names[ind] for ind in experiment_index]
n_subfigs = len(experiments)
corrected_p_threshold = args.alpha / sum(range(len(experiments)))
#Bonferroni correction
def fit_func(B, x):
"""y = m*x + b"""
return B[0] * x + B[1]
linear = Model(fit_func)
fig = plt.figure()
color_generator = iter(cm.magma(arange(0, 1, 1.0 / n_subfigs)))
z = 0
for i in range(n_subfigs):
experiment1 = experiments[i]
for j in range(n_subfigs):
experiment2 = experiments[j]
z += 1
vals_1 = {}
for line in data_lines:
atoms = line.split()
if atoms[experiment_index[i]] != 'NA':
#exclude NA values generated by ParTIES
score = float(atoms[experiment_index[i]])
ctrl_mean = mean([
float(atoms[n]) for n in range(len(atoms))
if n in args.control_columns
])
vals_1[atoms[0]] = max(score - ctrl_mean, 0)
vals_2 = {}
for line in data_lines:
atoms = line.split()
if atoms[experiment_index[j]] != 'NA':
#exclude NA values generated by ParTIES
score = float(atoms[experiment_index[j]])
ctrl_mean = mean([
float(atoms[n]) for n in range(len(atoms))
if n in args.control_columns
])
vals_2[atoms[0]] = max(score - ctrl_mean, 0)
common_keys = set(vals_1.keys()) & set(vals_2.keys())
x, y = [], []
for akey in common_keys:
x.append(vals_2[akey])
y.append(vals_1[akey])
slope, intercept, r_value, p_value, std_err = stats.linregress(
x, y)
rho, rho_p = stats.spearmanr(x, y)
if i > j:
if len(x) > 0 and len(y) > 0:
plt.subplot(n_subfigs, n_subfigs, z)
plt.hexbin(x,
y,
vmin=args.hexbin_vmin,
vmax=args.hexbin_vmax,
mincnt=1,
bins='log',
cmap=cm.viridis)
max_max = max(max(x), max(y))
xi = arange(0, max_max, 0.01)
line = slope * xi + intercept
mydata = RealData(array(x), array(y), sx=std(x), sy=std(y))
if not args.deactivate_ODR:
myodr = ODR(mydata, linear, beta0=[slope, intercept])
ODR_out = myodr.run()
if args.verbose:
ODR_out.pprint()
ODR_fitted_line = fit_func(ODR_out.beta, array(x))
lowess_arr = lowess(y, x, delta=0.01, return_sorted=True)
ys = lowess_arr[:, 1]
xs = lowess_arr[:, 0]
plt.plot(xs, ys, 'orange', lw=args.regression_line_width)
plt.plot(xi,
line,
ls='-',
color='red',
lw=args.regression_line_width)
if not args.deactivate_ODR:
plt.plot(x,
ODR_fitted_line,
"--",
color='gray',
lw=args.regression_line_width)
plt.axis(xmax=1.01, ymax=1.01, ymin=-0.01, xmin=-0.01)
ax = plt.gca()
current_xticklabels = ax.get_xticks()
current_yticklabels = ax.get_yticks()
ax.set_xticklabels([])
ax.set_yticklabels([])
plt.tick_params(axis='both',
which='both',
top='off',
right='off')
if j == 0:
ax.set_yticklabels(current_yticklabels,
fontsize=tick_font_size)
if args.show_axis_labels:
ax.set_ylabel(experiment1,
fontsize=label_font_size,
rotation=90,
style=args.font_style)
if i == n_subfigs - 1:
ax.set_xticklabels(current_xticklabels,
fontsize=tick_font_size,
rotation=45)
if args.show_axis_labels:
ax.set_xlabel(experiment2,
fontsize=label_font_size,
style=args.font_style)
elif i == j:
plt.subplot(n_subfigs, n_subfigs, z)
plt.tick_params(axis='both',
which='both',
top='off',
right='off')
if len(x) > 0:
plt.hist(x,
bins=arange(0, 1, 0.025),
linewidth=0.3,
ec="white",
fc=next(color_generator))
plt.axis(ymin=0, ymax=args.histogram_max)
ax = plt.gca()
ax.set_yticks(ax.get_yticks()[::2])
ax.set_xticks(arange(0, 1.2, 0.2))
current_xticklabels = ax.get_xticks()
current_yticklabels = ax.get_yticks()
ax.set_xticklabels([], fontsize=tick_font_size)
ax.set_yticklabels(current_yticklabels,
fontsize=tick_font_size)
if i == n_subfigs - 1:
ax.set_xticklabels(current_xticklabels,
fontsize=tick_font_size,
rotation=45)
if i > 0:
plt.tick_params(axis='both', labelleft='off')
else:
plt.subplot(n_subfigs, n_subfigs, z)
ax = plt.gca()
current_xticklabels = ax.get_xticks()
ax.set_xticklabels(current_xticklabels,
fontsize=tick_font_size)
if not args.use_pearson:
ax.set_facecolor(cm.YlGnBu(rho))
if rho > 0.5:
if rho_p < corrected_p_threshold:
plt.text(0.2,
0.4,
"%.2f" % round(rho, 3),
fontsize=1.5 * label_font_size,
color='white')
else:
plt.text(0.2,
0.4,
"%.2f^" % round(rho, 3),
fontsize=1.5 * label_font_size,
color='white')
else:
if rho_p < corrected_p_threshold:
plt.text(0.2,
0.4,
"%.2f" % round(rho, 3),
fontsize=1.5 * label_font_size,
color='black')
else:
plt.text(0.2,
0.4,
"%.2f^" % round(rho, 3),
fontsize=1.5 * label_font_size,
color='black')
else:
ax.set_facecolor(cm.YlGnBu(r_value))
if r_value > 0.5:
plt.text(0.2,
0.4,
"%.2f" % round(r_value, 3),
fontsize=1.5 * label_font_size,
color='white')
else:
plt.text(0.2,
0.4,
"%.2f" % round(r_value, 3),
fontsize=1.5 * label_font_size,
color='black')
plt.tick_params(axis='both',
which='both',
bottom='off',
top='off',
left='off',
right='off',
labelbottom='off',
labelleft='off',
labelright='off')
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.spines['bottom'].set_visible(False)
ax.spines['left'].set_visible(False)
if args.show_colorbar:
fig.subplots_adjust(right=0.8)
cbar_ax = fig.add_axes([0.85, 0.4, 0.015, 0.4])
cb_label = 'correlation plot hexbin scale (log10)'
cb = plt.colorbar(cax=cbar_ax,
drawedges=False,
ticks=range(args.hexbin_vmin, args.hexbin_vmax + 1),
label=cb_label)
current_yticklabels = cbar_ax.get_yticks()
cbar_ax.set_yticklabels(range(args.hexbin_vmin, args.hexbin_vmax + 1),
fontsize=tick_font_size)
cbar_ax.set_ylabel(cb_label, fontsize=label_font_size)
cb.outline.set_linewidth(1)
#cb.ax.get_children()[4].set_linewidths(1)
plt.savefig(args.output_matrix_image, dpi=((args.image_resolution)))
# d01 box
for i, (xlim, ylim, s,c) in enumerate(
zip([xlim_d01, xlim_d02, xlim_d03], [ylim_d01, ylim_d02, ylim_d03],
[15, 12, 9], ['b','g','r'])):
ax.add_patch(mpl.patches.Rectangle((xlim[0], ylim[0]),
xlim[1]-xlim[0], ylim[1]-ylim[0],
fill=None, lw=1.5, edgecolor='r', zorder=10))
ax.add_feature(OCEAN, lw=0.5, ec='k', zorder=0, facecolor='w')
ax.add_feature(LAND, lw=0.5, ec='k', zorder=1, facecolor='cornsilk')
ax.add_feature(LAKES, lw=0.5, ec='k', zorder=2, facecolor='w')
#ax.coastlines('50m', linewidth=1.5, )
white = np.ones((10,4))
high = cm.YlGnBu(np.linspace(0, 1, 128))
colors = np.vstack((white, high))
my_color = LinearSegmentedColormap.from_list('my_colormap',colors, N=24)
#
CS = ax.contour(to_np(lons), to_np(lats), roms.h,
levels=[10,50,100,200,500], linewidths=1.25, colors='k', transform=proj)
CS.levels = [nf(val) for val in CS.levels]
ax.clabel(CS, CS.levels, fmt='%r', inline=True, fontsize=9, )
cn = ax.contourf(to_np(lons), to_np(lats), roms.h, np.arange(0,7000,500), cmap=my_color, transform=proj, )
cbar = fig.colorbar(cn, ax=ax, shrink=0.6, )
cbar.set_label("ROMS Bathymetry (m)", rotation=-90, va='bottom', fontsize=14)
cbar.add_lines(CS)
lonmin, lonmax, latmin, latmax = [118, 128, 27, 41]
