def plot_matplotlib(sample, base_name):
colors = ['#e41a1c', '#377eb8', '#4daf4a', '#984ea3', '#ff7f00']
params = { 'figure.facecolor': 'white',
'figure.subplot.bottom': 0.0,
'font.size': 16,
'legend.fontsize': 16,
'legend.borderpad': 0.2,
'legend.labelspacing': 0.2,
'legend.handlelength': 1.5,
'legend.handletextpad': 0.4,
'legend.borderaxespad': 0.2,
'lines.markeredgewidth': 2.0,
'lines.linewidth': 2.0,
'axes.prop_cycle': plt.cycler('color',colors)}
plt.rcParams.update(params)
fig, ax = plt.subplots(2)
ax[0].hist(sample[met_pred_parameter], 50, range=(0.0, 400.0), label='Pred', histtype='step')
ax[0].hist(sample[met_reco_parameter], 50, range=(0.0, 400.0), label='Obs', histtype='step')
ax[0].hist(sample[met_target_parameter], 50, range=(0.0, 400.0), label='Target', histtype='step')
ax[0].set_xlim([0,400])
ax[0].set_xlabel("pT (GeV)")
ax[0].legend(loc='upper right')
ax[1].hist(sample[met_pred_resolution], 50, range=(-100.0, 100.0), label='Pred', histtype='step')
ax[1].hist(sample[met_reco_resolution], 50, range=(-100.0, 100.0), label='Obs', histtype='step')
ax[1].set_xlim([-100,100])
ax[1].set_xlabel("Resolution (%)")
ax[1].legend(loc='upper left')
fig.tight_layout(pad=0.3)
fig.savefig('{}.pdf'.format(base_name))
plt.show()
def test_get_color_cycle(self):
if mpl_ge_150:
colors = [(1., 0., 0.), (0, 1., 0.)]
prop_cycle = plt.cycler(color=colors)
with plt.rc_context({"axes.prop_cycle": prop_cycle}):
result = utils.get_color_cycle()
assert result == colors
def generate_feature_graph_for_repo(repo_count):
plt.rc('axes', prop_cycle=(plt.cycler('color', ['r', 'g', 'b', 'k']) + plt.cycler('linestyle', ['--', '--', '--', '--'])))
for per in xrange(1,5):
data = None
data = pd.read_csv('./features/repo_'+str(repo_count)+'_person_'+str(per)+'_uneven_person_commits.csv', sep=',', )
max_bins = data['Week_No'].max()
plt.plot(data['Week_No'], data['Commit_Count'], label="Person" + str(per))
plt.ylabel('Commit Count Per Person')
plt.xlabel('Week Number')
plt.title('Early Commit Smoke Per Person for Repo ' + str(repo_count))
plt.legend(bbox_to_anchor=(0., 1.02, 1., .102), loc=1, ncol=4, mode="expand", borderaxespad=0.)
plt.savefig('./results/repo_'+str(repo_count)+'_early_commit_smoke.png')
plt.close()
def setup(dpi=300, sketch=(1, 100, 2), theme='light'):
"""Setup travelmaps."""
# Customized plt.xkcd()-settings
# http://jakevdp.github.io/blog/2013/07/10/XKCD-plots-in-matplotlib
rcParams['font.family'] = ['Humor Sans', 'Comic Sans MS']
rcParams['font.size'] = 8.0
rcParams['path.sketch'] = sketch
rcParams['axes.linewidth'] = 1.0
rcParams['lines.linewidth'] = 1.0
rcParams['grid.linewidth'] = 0.0
rcParams['axes.unicode_minus'] = False
if theme=='dark':
rcParams['path.effects'] = [patheffects.withStroke(linewidth=2, foreground="k")]
rcParams['figure.facecolor'] = 'black'
rcParams['figure.edgecolor'] = 'black'
rcParams['lines.color'] = 'white'
rcParams['patch.edgecolor'] = 'white'
rcParams['text.color'] = 'white'
rcParams['axes.facecolor'] = 'black'
rcParams['axes.edgecolor'] = 'white'
rcParams['axes.labelcolor'] = 'white'
rcParams['xtick.color'] = 'white'
rcParams['ytick.color'] = 'white'
rcParams['grid.color'] = 'white'
rcParams['savefig.facecolor'] = 'black'
rcParams['savefig.edgecolor'] = 'black'
else:
rcParams['path.effects'] = [patheffects.withStroke(linewidth=2, foreground="w")]
rcParams['figure.facecolor'] = 'white'
rcParams['figure.edgecolor'] = 'white'
rcParams['lines.color'] = 'black'
rcParams['patch.edgecolor'] = 'black'
rcParams['text.color'] = 'black'
rcParams['axes.facecolor'] = 'white'
rcParams['axes.edgecolor'] = 'black'
rcParams['axes.labelcolor'] = 'black'
rcParams['xtick.color'] = 'black'
rcParams['ytick.color'] = 'black'
rcParams['grid.color'] = 'black'
rcParams['savefig.facecolor'] = 'white'
rcParams['savefig.edgecolor'] = 'white'
# *Bayesian Methods for Hackers*-colour-cylce
# (https://github.com/pkgpl/PythonProcessing/blob/master/results/matplotlibrc.bmh.txt)
rcParams['axes.prop_cycle'] = plt.cycler('color', ['#348ABD', '#A60628', '#7A68A6', '#467821', '#D55E00',
'#CC79A7', '#56B4E9', '#009E73', '#F0E442', '#0072B2'])
# Adjust dpi, so figure on screen and savefig looks the same
rcParams['figure.dpi'] = dpi
rcParams['savefig.dpi'] = dpi
def set_palette(name=None, num_colors=8, start=0):
"""Set the active color palette
Parameters
----------
name : str, optional
Name of the palette. If `None`, modify the active palette.
num_colors : int
Number of colors to retrieve.
start : int
Staring from this color number.
"""
palette = get_palette(name, num_colors, start)
mpl.rcParams["axes.prop_cycle"] = plt.cycler('color', palette)
mpl.rcParams["patch.facecolor"] = palette[0]
def plot_energies_on_plane(folder_loc):
"""
Plot Poincare 3D surface on a 2D plane with the colour of the plane defined
by the energy of that point.
"""
cm = plt.cm.viridis
plt.gca().set_prop_cycle(
plt.cycler('color', cm(np.linspace(0, 1, len(range(470))))))
t = [0.1 * i for i in range(1200)]
for i in range(1, 550):
try:
with open(folder_loc + f"results_{i}.yml") as f:
data = yaml.load(f, Loader=yaml.Loader)
except:
continue
result = data['data']
energies = apply_energy(result)
mod_θ2 = centre_plane(result[:, 0])
x_points = find_crossing_points(mod_θ2)
es = []
ϕ1s = []
ϕ2s = []
θ2s = []
for x in x_points:
f1, f2, f3, f4 = interpolate_functions(x[0], t, result)
t0 = find_zero(x[0], t, f1)
ϕ1_0 = f2(t0)
ϕ2_0 = f3(t0)
θ2_0 = f4(t0)
es.append(energies[x[0]])
ϕ1s.append(ϕ1_0)
ϕ2s.append(ϕ2_0)
θ2s.append(θ2_0)
plt.scatter(np.mod(np.array(ϕ1s), np.pi * 2),
np.mod(np.array(θ2s), np.pi * 2),
c=es,
s=0.5)
plt.colorbar()
plt.show()
def plot_displacement_all(filenames):
fig = plt.figure(figsize=(6, 6), dpi=80, facecolor='w', edgecolor='k')
ax = fig.add_subplot(111)
displacement = np.loadtxt(filenames[0], delimiter=',')
plt.set_cmap(plt.get_cmap('winter'))
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.jet(np.linspace(0.1, 0.9, len(filenames)))))
xmax = 0
for filename in filenames:
xmax = max(xmax, np.loadtxt(filename, delimiter=',')[:, 0][-1])
ax.set_xlim(left=0.01, right=xmax)
# ax.set_ylim(bottom=100, top=10000000)
plt.xlabel('$t [s]$')
plt.ylabel('$< \Delta r^2 > [\mu m^2]$')
plt.yscale('log')
plt.xscale('log')
for filename in filenames:
displacement = np.loadtxt(filename, delimiter=',')
time = list(displacement[:, 0][:100])
disp = list(displacement[:, 1][:100])
i = 0
while 101 + int(pow(1.05, i)) < len(displacement[:, 0]):
time.append(displacement[:, 0][101 + int(pow(1.05, i))])
disp.append(
sum(displacement[:, 1][100 + int(pow(1.05, (i - 1))):101 +
int(pow(1.05, i))]) /
(1 + int(pow(1.05, i)) - int(pow(1.05, (i - 1)))))
i += 1
time.append(displacement[:, 0][-1])
disp.append(
sum(displacement[:, 1][100 + int(pow(1.05, (i - 1))):]) /
(len(displacement[:, 1]) - int(pow(1.05, (i - 1)))))
plt.plot(time, disp, alpha=0.7, label=filename[7:].split('*')[0])
plt.legend()
filename = filename.split(' = ')[0] + filename[:-4].split(
' * ')[1] + '.png'
filename = filename.replace('\\', '')
filename = filename.replace('$', '')
filename = filename.replace('{', '')
filename = filename.replace('}', '')
filename = filename.replace(' ', '')
plt.savefig(filename, bbox_inches='tight')
def visualize_results(file_name, title_name, graph_type, type_weights,
fin_normalized_load, fin_approx_h1, fin_approx_h2,
fin_h1, fin_h2):
fig, ax = plt.subplots()
ax.set_prop_cycle(
plt.cycler('color', ['black', 'saddlebrown', '#ff3232', 'gold']))
ax.plot(fin_normalized_load,
fin_approx_h1,
'v-',
linewidth=1.7,
label='local-ratio ' + r'$h_1$')
ax.plot(fin_normalized_load,
fin_h1,
's--',
linewidth=1.5,
dashes=[5, 5, 5, 5],
label=r'$h_1$')
ax.plot(fin_normalized_load,
fin_approx_h2,
'o-',
linewidth=1.7,
label='local-ratio ' + r'$h_1$')
ax.plot(fin_normalized_load,
fin_h2,
'd--',
linewidth=1.5,
dashes=[5, 5, 5, 5],
label=r'$h_2$')
ax.legend(loc='upper left')
ax.set_xlabel('normalized load')
ax.set_ylabel(title_name)
ax.set_xticklabels(labels=[], minor=True)
ax.set_title(title_name)
ax.set_yscale("log")
plt.draw()
plt.pause(5)
plt.savefig('plots/' + graph_type + "/" + type_weights + "/" + file_name +
'.pdf')
def plot_density_all(filenames):
fig = plt.figure(figsize=(8, 4), dpi=80, facecolor='w', edgecolor='k')
ax = fig.add_subplot(111)
plt.xlabel('')
plt.ylabel('')
plt.set_cmap(plt.get_cmap('winter'))
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.jet(np.linspace(0.1, 0.9, len(filenames)))))
for filename in filenames:
concentration = np.loadtxt(filename, delimiter=',')
plt.plot(concentration, alpha=0.7, label=filename[7:].split('*')[0])
plt.legend()
filename = filename.split(' = ')[0] + filename[:-4].split(
' * ')[1] + '.png'
filename = filename.replace('\\', '')
filename = filename.replace('$', '')
filename = filename.replace('{', '')
filename = filename.replace('}', '')
filename = filename.replace(' ', '')
plt.savefig(filename, bbox_inches='tight')
def set_palette(*args, **kwds):
"""Set the matplotlib color cycler.
args, kwds: same as for sns.color_palette
Also takes a boolean kwd, `reverse`, to indicate
whether the order of the palette should be reversed.
returns: list of colors
"""
reverse = kwds.pop('reverse', False)
palette = sns.color_palette(*args, **kwds)
palette = list(palette)
if reverse:
palette.reverse()
cycler = plt.cycler(color=palette)
plt.gca().set_prop_cycle(cycler)
return palette
def set_style(style='blackviz'):
plt.rc('figure', facecolor='0.065', autolayout=False, dpi=120)
plt.rc('axes',
facecolor='0.125',
edgecolor='w',
grid=True,
labelcolor='lightgrey',
labelsize='large',
prop_cycle=plt.cycler('color', [
'deepskyblue', 'orangered', 'green', 'yellow', 'mintcream',
'magenta', 'darkorange'
]))
plt.rc('xtick', color='w')
plt.rc('ytick', color='w')
plt.rc('grid', color='w', linestyle='-', alpha=0.25)
plt.rc('font', size=12)
plt.rc('text', color='lightgrey')
plt.rc('lines', linewidth=2)
plt.rc('figure', figsize='10, 8', titlesize='x-large')
def set_draw_settings(self):
params = {
"figure.facecolor":
"#cad9e1",
"axes.facecolor":
"#cad9e1",
"axes.grid":
True,
"axes.grid.axis":
"y",
"grid.color":
"#ffffff",
"grid.linewidth":
2,
"axes.spines.left":
False,
"axes.spines.right":
False,
"axes.spines.top":
False,
"ytick.major.size":
0,
"ytick.minor.size":
0,
"xtick.direction":
"in",
"axes.titlesize":
"small",
"xtick.major.size":
7,
"xtick.color":
"#191919",
"axes.edgecolor":
"#191919",
"axes.prop_cycle":
plt.cycler('color', [
'#006767', '#ff7f0e', '#2ca02c', '#d62728', '#9467bd',
'#8c564b', '#e377c2', '#7f7f7f', '#bcbd22', '#17becf'
])
}
plt.rcParams.update(params)
def rmse_time(al_rmse, avg_sim_time, n_node, hop_lim, n_trial, results_path):
plt.style.use('bmh')
plt.rcParams['axes.prop_cycle'] = plt.cycler(color=plt.cm.Paired.colors)
plt.rcParams['font.sans-serif'] = 'Arial'
fig, ax = plt.subplots(figsize=(16, 8))
x = range(1, n_trial + 1)
# all_rmse = [1000*i for i in al_rmse]
x_ticks = list(range(0, n_trial + 1, 10))
ticks = [1, '', '', '', '', '', '', '', '', '']
for i in range(1, -(-n_trial // 10)):
sub_ticks = [i * 10, '', '', '', '', '', '', '', '', '']
ticks += sub_ticks
ticks = ticks[:n_trial]
ax.plot(x, al_rmse, color='C1', linewidth=1)
ax.set_xlabel('Trial #', fontsize=11)
ax.set_ylabel("Average RMSE [km]", fontsize=11)
ax.grid(axis='x')
ax.set_axisbelow(True)
ax.set_title(f'{n_node} Node Network RMSE', {'fontsize': 13})
ax.set_xlim(0, n_trial + 1)
ax.set_ylim(min(al_rmse) * .5, max(al_rmse) * 1.1)
ax.set_xticks(x_ticks)
# ax.xaxis.set_ticklabels([str(i) for i in list(x)])
mean = r'$\mu_{rmse} = $' + f"{np.mean(al_rmse):.2f} km" + '\n' + \
r'$\mu_{sim time} = $' + f"{avg_sim_time:.2f} s"
ax.text(.98,
.90,
mean,
transform=ax.transAxes,
va='baseline',
fontsize=11,
ha='right',
bbox=dict(facecolor='C0', edgecolor='k', alpha=.6))
fig.savefig(f'{results_path}/{hop_lim}hop_{n_trial}sim_rmse-refine.png',
dpi=200)
plt.show()
def _make_style():
nearly_black = '0.15'
linewidth = 0.6
dpi = 160
palette = list(get_palette('Set1'))
palette[5] = list(get_palette('Set2'))[5]
defaults = mpl_style.library['classic']
style = {
'lines.solid_capstyle': 'round', # [projecting] butt|round|projecting
'font.size': 7.0, # [12.0]
'text.color': nearly_black, # [black]
'mathtext.default': 'regular', # [it] the default font to use for math.
'axes.edgecolor': nearly_black, # [black] axes edge color
'axes.linewidth': linewidth, # [1.0] edge linewidth
'axes.labelcolor': nearly_black, # [black]
'axes.unicode_minus': False, # [True] use unicode for the minus symbol
'axes.prop_cycle': plt.cycler('color', palette), # ['bgrcmyk']
'patch.facecolor': palette[1], # [b]
'xtick.major.size': 2.5, # [4] major tick size in points
'xtick.minor.size': 1.0, # [2] minor tick size in points
'xtick.major.width': linewidth, # [0.5] major tick width in points
'xtick.color': nearly_black, # [black] color of the tick labels
'ytick.major.size': 2.5, # [4] major tick size in points
'ytick.minor.size': 1.0, # [2] minor tick size in points
'ytick.major.width': linewidth, # [0.5] major tick width in points
'ytick.color': nearly_black, # [black] color of the tick labels
'legend.fancybox': True, # [False] Use a rounded box for the legend
'legend.numpoints': 1, # [2] the number of points in the legend line
'legend.fontsize': 'medium', # ['large']
'legend.framealpha': 0.9, # [None] opacity of of legend frame
'figure.figsize': (3.4, 2.8), # [(8, 6) inch] (3.4, 2.8) inch == (8.6, 7.1) cm
'figure.dpi': dpi, # [80] figure dots per inch
'figure.facecolor': 'white', # [0.75] figure facecolor
'image.cmap': 'viridis', # [jet...]
'savefig.dpi': dpi, # [100] figure dots per inch
'savefig.bbox': 'tight', # ['standard']
'savefig.pad_inches': 0.04, # [0.1] padding to be used when bbox is set to 'tight'
}
return with_defaults(style, defaults)
def plot_insulation(clr, insulation, windows, resolution, out_path, exclude_chroms, title):
dir_path = os.path.join(os.path.dirname(out_path), title)
if not os.path.exists(dir_path):
os.mkdir(dir_path)
chromsizes = bioframe.fetch_chromsizes('sacCer3', filter_chroms=False)
regions = [(k, 0, v) for k, v in chromsizes.drop('chrM').iteritems()]
for region in regions:
norm = LogNorm(vmax=0.1, vmin=0.001)
data = clr.matrix(balance=True).fetch(region)
fig, ax = plt.subplots(figsize=(20, 4))
img = plot_45_mat(ax, data, start=0, resolution=resolution, norm=norm, cmap='fall')
ax.set_aspect(0.5)
ax.set_ylim(0, 30000)
format_ticks(ax, rotate=False)
ax.xaxis.set_visible(False)
divider = make_axes_locatable(ax)
cax = divider.append_axes('right', size='1%' ,pad=0.1, aspect=6)
plt.colorbar(img, cax=cax)
insul_region = bioframe.select(insulation, region)
ins_ax = divider.append_axes('bottom', size='50%', pad=0.0, sharex=ax)
ins_ax.set_prop_cycle(plt.cycler('color', plt.cm.plasma(np.linspace(0, 1, 5))))
for window in windows:
ins_ax.plot(insul_region[['start', 'end']].mean(axis=1),
insul_region[f'log2_insulation_score_{window}'],
label=f'{window} bp window', lw=1)
ins_ax.legend(bbox_to_anchor=(1.125, 1.05), loc='upper right')
fig.suptitle(f'{title}: {region[0]}')
path = os.path.join(dir_path, '_'.join((region[0], os.path.basename(out_path))))
plt.savefig(path, dpi=300)
def time_plot2(df, sample_name, labels):
# Initialise figure
fig, ax = plt.subplots(figsize=(16, 4))
ax.plot(df)
# Define colour space for lines
colourmap = plt.cm.gist_ncar
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.jet(np.linspace(0, 0.9, 9))))
# Set up layout
ax.set(xlabel='time steps',
ylabel='signal (analog counts)',
title='data sample: ' + sample_name)
ax.legend(labels)
ax.grid()
# fig.savefig("data/figures/"+sample_name+"_timeplot.png")
plt.show()
return
def plot(self, var_to_plot=None):
if var_to_plot is None:
var_to_plot = self.labels
fig = plt.figure()
ax = fig.add_subplot(111)
ax.set_prop_cycle(
plt.cycler('color', ['r', 'g', 'b', 'c', 'm', 'y', 'k']))
handles = []
labels = []
for v in var_to_plot:
try:
i = self.labels.index(v)
h, = ax.plot(self.time, self.data[:, i])
handles.append(h)
labels.append(v)
except:
print("Variable", v, "not found")
fig.legend(handles, labels)
plt.title(self.description)
plt.xlabel('Time')
plt.show()
def usePlotStyle():
matplotlib.style.use('default')
plt.rcParams.update({
'figure.figsize': (20, 10),
'font.size': 20,
'xtick.labelsize': 18,
'ytick.labelsize': 18,
'axes.prop_cycle': plt.cycler(
color=[
Colors.orange,
Colors.blue,
Colors.green,
Colors.red,
Colors.purple,
Colors.brown,
Colors.pink,
Colors.gray,
Colors.yellow,
Colors.turquoise
])
})
def cycle(self, n: str):
if n == '0':
return plt.cycler("color", self.schemes(
self.parameters['scheme'])) #colors)
elif n == '1':
return plt.cycler("color", self.schemes(
self.parameters['scheme'])) + plt.cycler(
"linestyle", self.linestyles(self.parameters['linestyle']))
elif n == '2':
return plt.cycler("color", self.schemes(
self.parameters['scheme'])) * plt.cycler(
"linestyle", self.linestyles(self.parameters['linestyle']))
elif n == '3':
return plt.cycler(
"linestyle", self.linestyles(
self.parameters['linestyle'])) * plt.cycler(
"color", self.schemes(self.parameters['scheme']))
else:
raise ValueError('Index out of bound')
def plotExTimes(execTimesDict):
'''
Given a dict funcname:executiontimesarray
plots all the functions in the same plot
'''
# Make a colormap to plot the lines in different colors
# pylint: disable=no-member,unused-variable
num_plots = len(execTimesDict)
fig, ax = plt.subplots(figsize=(15, 3))
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.gist_rainbow(np.linspace(0, 1, num_plots))))
for fname, exTime in execTimesDict.items():
ax.plot(exTime, label=fname)
ax.set_xticklabels([str(i / 5.) for i in range(6)])
ax.set_xticks([i for i in range(0, 101, 20)])
ax.set_xlabel('u')
ax.set_ylabel('Execution time')
ax.legend(loc='best')
plt.suptitle('Runtime execution times')
def setup_noxkcd(dpi=300, theme='light'):
"""Setup Maps."""
if theme=='dark':
rcParams['figure.facecolor'] = 'black'
rcParams['figure.edgecolor'] = 'black'
rcParams['lines.color'] = 'white'
rcParams['patch.edgecolor'] = 'white'
rcParams['text.color'] = 'white'
rcParams['axes.facecolor'] = 'black'
rcParams['axes.edgecolor'] = 'white'
rcParams['axes.labelcolor'] = 'white'
rcParams['xtick.color'] = 'white'
rcParams['ytick.color'] = 'white'
rcParams['grid.color'] = 'white'
rcParams['savefig.facecolor'] = 'black'
rcParams['savefig.edgecolor'] = 'black'
else:
rcParams['figure.facecolor'] = 'white'
rcParams['figure.edgecolor'] = 'white'
rcParams['lines.color'] = 'black'
rcParams['patch.edgecolor'] = 'black'
rcParams['text.color'] = 'black'
rcParams['axes.facecolor'] = 'white'
rcParams['axes.edgecolor'] = 'black'
rcParams['axes.labelcolor'] = 'black'
rcParams['xtick.color'] = 'black'
rcParams['ytick.color'] = 'black'
rcParams['grid.color'] = 'black'
rcParams['savefig.facecolor'] = 'white'
rcParams['savefig.edgecolor'] = 'white'
# *Bayesian Methods for Hackers*-colour-cylce
# (https://github.com/pkgpl/PythonProcessing/blob/master/results/matplotlibrc.bmh.txt)
rcParams['axes.prop_cycle'] = plt.cycler('color', ['#348ABD', '#A60628', '#7A68A6', '#467821', '#D55E00',
'#CC79A7', '#56B4E9', '#009E73', '#F0E442', '#0072B2'])
# Adjust dpi, so figure on screen and savefig looks the same
rcParams['figure.dpi'] = dpi
rcParams['savefig.dpi'] = dpi
def transmittance(date_dir, atmfile, filters, plot=True):
"""
"""
# Read atmfile
molecules, p, T, abundances = mat.readatm(date_dir + atmfile)
nlayers = len(p)
p = p[::-1] # top to bottom of the atmosphere
# Read tau.dat
foo = date_dir + 'tau.dat'
tau, wns = readTauDat(foo, nlayers)
# Transmittance:
transmit = np.exp(-tau)
# Band intgrate it:
filt_tr = filter_cf(filters, nlayers, wns, transmit)
nfilters = len(filters)
colors = plt.cm.rainbow(np.asarray(np.linspace(0, 255, nfilters), np.int))
if plot:
print(" Plotting contribution functions.\n")
# Not normalized cf
plt.figure(4)
plt.clf()
gs = gridspec.GridSpec(1, 2, width_ratios=[5, 1])
ax0 = plt.subplot(gs[0])
colormap = plt.cm.rainbow(np.linspace(0, 1, len(filters)))
ax0.set_prop_cycle(plt.cycler('color', colormap))
for i in np.arange(len(filt_tr)):
(head, tail) = os.path.split(filters[i])
lbl = tail[:-4]
ax0.semilogy(filt_tr[i], p, '-', linewidth = 1.5, label=lbl,
color=colors[i])
lgd = ax0.legend(loc='center left', bbox_to_anchor=(1.0, 0.5),
ncol=nfilters//30 + 1, prop={'size':8})
ax0.set_ylim(max(p), min(p))
ax0.ticklabel_format(style='sci', axis='x', scilimits=(0,0))
ax0.set_xlabel('Transmittance', fontsize=14)
ax0.set_ylabel('Pressure (bar)' , fontsize=14)
plt.savefig(date_dir + 'Transmittance.png')
return filt_tr[:,::-1]
def set_defaults():
"""
Set default plotting behavior.
"""
start_colors = ["#D11700", "#8EBA42", "#348ABD", "#988ED5", "#777777",
"#FBC15E", "#FFB5B8"]
if LooseVersion(mpl.__version__) < LooseVersion("1.5.1"):
plt.rcParams["axes.color_cycle"] = start_colors
else:
plt.rcParams["axes.prop_cycle"] = plt.cycler("color", start_colors)
plt.rcParams["axes.facecolor"] = "w"
plt.rcParams["axes.grid"] = True
plt.rcParams["axes.axisbelow"] = True
plt.rcParams["axes.linewidth"] = 0
plt.rcParams["axes.labelcolor"] = "k"
plt.rcParams["figure.facecolor"] = "w"
plt.rcParams["figure.subplot.bottom"] = 0.125
plt.rcParams["figure.subplot.left"] = 0.1
plt.rcParams["lines.linewidth"] = 2
plt.rcParams["grid.color"] = "#444444"
plt.rcParams["grid.linewidth"] = 1.5
plt.rcParams["grid.linestyle"] = ":"
plt.rcParams["xtick.direction"] = "out"
plt.rcParams["ytick.direction"] = "out"
plt.rcParams["xtick.color"] = "k"
plt.rcParams["ytick.color"] = "k"
plt.rcParams["text.color"] = "k"
plt.rcParams["image.interpolation"] = "nearest"
plt.rcParams["image.cmap"] = "gist_gray"
return
def plot_trend(self, items):
"""
:param items: list of items we wish to plot
:return: plot trending items
"""
num_plots = len(items)
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.Accent(np.linspace(0, 1, num_plots))))
# colormap=plt.cm.gist_ncar
# plt.gca().set_color_cycle([colormap(i) for i in np.linspace(0, 0.9, num_plots)])
labels = []
for item in items:
df_filtered = self.data[self.data['item_id'] == item]
_lag = df_filtered.lag.tolist()[::-1]
_num_user = df_filtered.user_count.tolist()[::-1]
plt.plot(_lag, _num_user)
labels.append(item)
plt.legend(labels, ncol=5, loc='upper center')
plt.title('Top Trending Items')
plt.ylabel('Number of transactions')
plt.xlabel('Month-end lag')
return plt.show()
def visualize_dummy_samples(x, y):
sns.set_style('whitegrid')
tsne_d = TSNE(init='pca')
plt.rcParams['axes.prop_cycle'] = plt.cycler(color=[
"#FF0000", "#005F0E", "#072480", "#55025B", "#FE7878", "#89FA82",
"#828FFA", "#F38BD8", "#616061"
])
# Dimensionality reduction on the embeddings using t-SNE
emb_d = tsne_d.fit_transform(x)
plt.figure(figsize=[10, 7])
labels = y.cpu().numpy()
for i in np.unique(labels):
class_ind = np.where(labels == i)
plt.scatter(emb_d[class_ind, 0],
emb_d[class_ind, 1],
label=f'{i}',
alpha=0.5,
s=5)
# plt.legend()
plt.xlabel('t-SNE component 1')
plt.ylabel('t-SNE component 2')
def plot_rewards(simulation_list, list_type="s") -> plt.Figure:
"""
Plots the cumulative reward of given simulations
:param simulation_list: list of Simulation objects, whose run_simulation() method was called
:param list_type str, s for simulation, d for data. s means simulation list contains Simulation objects,
d means simulation list contains tuples of label,data
:return: the figure with the plots
"""
global colormap, fontP
fontP.set_size('small' if len(simulation_list) > 5 else 'medium')
fig = plt.figure()
ax = fig.subplots()
ax.set_prop_cycle(
plt.cycler('color', colormap(np.linspace(0, 1, len(simulation_list)))))
for sim in simulation_list:
ax.plot(sim.get_reward_sum() if list_type == SIMULATION else sim[1],
label=sim.type if list_type == SIMULATION else sim[0])
ax.legend(prop=fontP, framealpha=FRAME_ALPHA)
ax.set_title("Cumulative Reward")
ax.set_xlabel(r"Trial")
ax.set_ylabel(r"Cumulative Reward")
return fig
def plotBestTeamsWithTupleData(environment_width, environment_height,
team_size, runs_to_average, max_steps,
min_sensor_radius, max_sensor_radius):
# Assumes files are spread out across multiple files with distinct data
test_11_data = []
for i in range(min_sensor_radius, max_sensor_radius + 1):
filename = "../../test11_%s_%s_%s_%s_%s_iter%s.p" % (
environment_width, environment_height, team_size, runs_to_average,
max_steps, i)
test_11_data += pickle.load(open(filename, "rb"))
# Plot Avg Completion Time vs Roomba Ratio
sensor_radii_to_plot = list(range(min_sensor_radius,
max_sensor_radius + 1))
plt.gca().set_prop_cycle(
plt.cycler('color',
plt.cm.jet(np.linspace(0, 0.9, len(sensor_radii_to_plot)))))
for j, sensor_radii in enumerate(sensor_radii_to_plot):
filtered_data = list(
filter(lambda tup: tup[0] == sensor_radii, test_11_data))
# Extract sensor radius, ratio, and completion time
s = [tup[0] for tup in filtered_data]
r = [tup[1] for tup in filtered_data]
c = [np.average(tup[2]) for tup in filtered_data]
# Annotate the minimum for min/max sensor radius
if sensor_radii == min_sensor_radius or sensor_radii == max_sensor_radius:
annot_min(np.array(r), np.array(c))
plt.plot(r, c)
legend = ["sensor_radius=" + str(x) for x in sensor_radii_to_plot]
plt.legend(legend, loc='upper left')
plt.title(
'Roomba/Drone Teams of Size %s for varying drone sensor_radius (%sx%s env, average over %s samples each)'
% (team_size, environment_height, environment_width, runs_to_average))
plt.xlabel('Ratio of Roomba')
plt.ylabel('Avg Completion Time')
plt.show()
def _plot(script_runs, min_support_range, show_graph):
script_names = list(script_runs.keys())
num_plots = len(script_names)
color_map = plt.cm.gist_ncar
plt.gca().set_prop_cycle(plt.cycler('color', plt.cm.jet(np.linspace(0, 1, num_plots))))
for script_name in script_names:
plt.plot(range(1, 21), script_runs[script_name])
plt.legend(script_names)
plt.ylim(ymin=0)
# plt.xlim(xmin=0)
plt.xticks(np.arange(21))
plt.xlabel('number of bins')
plt.ylabel('average top 10 lifts')
grpah_file_name = './img/{}.png'.format(datetime.datetime.now())
plt.savefig(grpah_file_name)
print('Graph saved in {}'.format(grpah_file_name))
if show_graph:
plt.show()
def update_colorcycler(N: int,
cmap: str = 'viridis',
ax: None or plt.Axes = None):
"""Update color cycler to better automate and choose the colors of certain things
Parameters
----------
N : int
Number of colors wanted to cycle through
cmap : str, optional
name of the colormap to cycle through, by default 'viridis'
ax : None or plt.Axes, optional
Optional Axes to update the color cycler of, by default None
See Also
--------
lwsspy.plot.pick_colors_from_cmap.pick_colors_from_cmap
Notes
-----
:Author:
Lucas Sawade ([email protected])
:Last Modified:
2020.01.13 14.30
"""
# Get Colors
colors = lpy.plot.pick_colors_from_cmap(N, cmap=cmap)
# Set axes
if ax is None:
ax = plt.gca()
# Set axes colorcycler
ax.set_prop_cycle(plt.cycler('color', colors))
def top_states(buffer: io.BytesIO, days: int) -> None:
plt.rcParams['axes.prop_cycle'] = plt.cycler(color=["#6886c5", "#f1935c", "#a1cae2", "#cd5d7d", "#838383"])
CSV_TIME_SERIES: str = "https://api.covid19india.org/csv/latest/state_wise_daily.csv"
df = pd.read_csv(CSV_TIME_SERIES)
df = df.drop(['Date', 'TT'], axis=1)
index = pd.date_range(start=df['Date_YMD'][0], end=df['Date_YMD'][len(df) - 1], freq="D")
index = [pd.to_datetime(date, format='%Y-%m-%d').date() for date in index]
confirmed = df.iloc[::3, :]
confirmed.index = index
if days:
confirmed = confirmed.tail(days)
confirmed = confirmed.dropna().drop(['Date_YMD', 'Status'], axis=1)
s = confirmed.sum().sort_values(ascending=False, inplace=False)
confirmed = confirmed[s.index[:5]]
confirmed: pd.DataFrame = confirmed[confirmed.select_dtypes(include=[np.number]).ge(0).all(1)]
for column in confirmed.columns.values:
confirmed.rename(columns={column: state_dict[column]}, inplace=True)
ax = confirmed.dropna().plot(kind='line', linewidth=2.0)
if not days or (days and days > 100):
ax.xaxis.set_major_locator(mdates.MonthLocator(interval=1))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%B %Y'))
plt.xlabel("Month")
plt.ylabel("Number of Cases")
plt.title("COVID-19 India - Top 5 Most Affected States")
plt.gcf().autofmt_xdate()
plt.savefig(buffer, format='png', bbox_inches='tight', pad_inches=0.1)
del [df, confirmed, s]
def plot_observables_allints(self, saveplot=True, odir="./"):
"""
Plot closure phases and visibility amplitudes vs. baseline length
for all integrations of an observable set.
Integrations different colors (really not useful)
"""
if self.multi:
t3_bl = self.geometry.t3_bl
all_cps = self.oi.OI_T3.T3PHI
vis_bl = self.geometry.vis_bl
all_visamps = self.oi.OI_VIS.VISAMP
nints = all_cps.shape[1]
colormap = plt.cm.gist_ncar
plt.gca().set_prop_cycle(
plt.cycler("color", plt.cm.jet(np.linspace(0, 1, nints)))
)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(20, 7))
for ii in np.arange(nints):
ax1.plot(t3_bl, all_cps[:, ii], ".")
ax2.plot(vis_bl, all_visamps[:, ii], ".")
ax1.set_xlabel(r"$B_{max}$", size=14)
ax1.set_ylabel("Closure phase [deg]", size=14)
ax1.set_title("Closure Phase", size=16)
ax2.set_title("Visibility Amplitude", size=16)
ax2.set_xlabel(r"$B_{max}$", size=14)
ax2.set_ylabel("Visibility Amplitude", size=14)
plt.suptitle(self.fn)
if saveplot:
plotname = os.path.join(odir, self.fn + "_observables_allints.png")
plt.savefig(plotname)
else:
plt.show()
else:
print(
"This is not a multi-integration OIFITS file, use plot_observables function instead."
)
def main(run_id="perf_shift", model_prefix="lenet5"):
N = 7
plt.rcParams["axes.prop_cycle"] = plt.cycler(
"color", plt.cm.viridis(np.linspace(0, 1, N)))
plt.rcParams.update({'font.size': 14})
ENABLE_SAVE_FIGURES = True
RESULTS_DIRECTORY = os.path.relpath(os.path.join(os.getcwd(), 'results'))
log.info("Results directory: %s", RESULTS_DIRECTORY)
RUN_ID = run_id
# set data folders
res_dir_list = list()
IMGS_PATH = os.path.join(RESULTS_DIRECTORY, RUN_ID)
res_dir_list = glob.glob(f"{IMGS_PATH}/{model_prefix}*")
log.info(f"Result directories: {res_dir_list}")
# PLOT
figures = dict()
# # shifted data
model_prefix = model_prefix.replace('-', '')
figures[f"ld-{model_prefix}-rotated.png"] = plot_rotated(res_dir_list)
figures[f"ld-{model_prefix}-shifted.png"] = plot_shifted(res_dir_list)
figures[f"ld-{model_prefix}-count60.png"] = plot_confidence_vs_count_60(
res_dir_list)
figures[f"ld-{model_prefix}-acc60.png"] = plot_confidence_vs_accuracy_60(
res_dir_list)
log.info("cwd: %s", os.getcwd())
if ENABLE_SAVE_FIGURES:
log.info("Saving plots")
# create save folder
log.info("image path: %s", IMGS_PATH)
os.makedirs(IMGS_PATH, exist_ok=True)
# save loop
for fn in figures:
figures[fn].savefig(os.path.join(IMGS_PATH, fn), dpi=200)
def create_plots(data: List[Tuple[Tuple[List[float], List[int], List[int], List[int]], str]]) -> None:
colors = ['#2CBDFE', '#47DBCD', '#F3A0F2', '#9D2EC5', '#661D98', '#F5B14C']
plt.rcParams['axes.prop_cycle'] = plt.cycler(color=colors)
heap_fig = plt.figure(num='matrace-analyzer (heap usage)')
stack_fig = plt.figure(num='matrace-analyzer (stack usage)')
anon_fig = plt.figure(num='matrace-analyzer (anon page usage)')
heap_axes = heap_fig.gca(title='Heap usage', xlabel='Time (sec)', ylabel='Memory (kilobytes)')
stack_axes = stack_fig.gca(title='Stack usage', xlabel='Time (sec)', ylabel='Memory (kilobytes)')
anon_axes = anon_fig.gca(title='Anon page usage', xlabel='Time (sec)', ylabel='Memory (kilobytes)')
for d in data:
(time, heap, stack, anon), filename = d
heap_axes.plot(time, heap, label=filename)
stack_axes.plot(time, stack, label=filename)
anon_axes.plot(time, anon, label=filename)
heap_axes.legend()
stack_axes.legend()
anon_axes.legend()
plt.show()
def plot_vol(f_vol, min_history=52):
"""
plot weekly factor volatilities
Inputs
----------
f_vol: dataframe
dataframe containing the weekly factor volatilities
min_history: int, Default = 52
minimum number of weeks to calculate volatility
Outputs
-------
None
"""
plt.figure(figsize=(12, 6))
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.jet(np.linspace(0, 1, f_vol.shape[1]))))
plt.plot(100 * f_vol[min_history:])
plt.legend(f_vol.columns)
plt.title('Weekly Factor Volatilities (%)')
def plot_map_by_number(angles, show=True, argsort=False):
cm = plt.cm.viridis
plt.gca().set_prop_cycle(
plt.cycler('color', cm(np.linspace(0, 1, len(angles)))))
for i, solution in tq.tqdm(enumerate(angles)):
t = solution.t
results = solution.y.T
mod_θ2 = centre_plane(results[:, 0])
x_points = find_crossing_points(mod_θ2)
ϕ1s = []
ϕ2s = []
θ2s = []
for x in x_points:
if x[0] <= 1: continue
f1, f2, f3, f4 = interpolate_functions(x[0], t, results)
t0 = find_zero(x[0], t, f1)
ϕ1_0 = f2(t0)
ϕ2_0 = f3(t0)
θ2_0 = f4(t0)
ϕ1s.append(ϕ1_0)
ϕ2s.append(ϕ2_0)
θ2s.append(θ2_0)
if argsort:
phi1_index = np.mod(np.array(ϕ1s), 2 * np.pi).argsort()
plt.plot(np.mod(np.array(ϕ1s)[phi1_index], np.pi * 2),
np.mod(np.array(θ2s)[phi1_index], np.pi * 2),
ms=0.5,
marker='o')
else:
plt.scatter(np.mod(np.array(ϕ1s), np.pi * 2),
np.mod(np.array(θ2s), np.pi * 2),
s=0.5)
if show:
plt.colorbar()
plt.show()
def total(buffer: io.BytesIO, days: int) -> None:
plt.rcParams['axes.prop_cycle'] = plt.cycler(
color=['#ea5455', '#1fab89', '#b19cd9'])
CSV_TIME_SERIES: str = "https://api.covid19india.org/csv/latest/case_time_series.csv"
df = pd.read_csv(CSV_TIME_SERIES)
df = df.drop(
['Total Confirmed', 'Total Recovered', 'Total Deceased', 'Date'],
axis=1,
errors='ignore')
index = pd.date_range(start=df['Date_YMD'][0],
end=df['Date_YMD'][len(df) - 1],
freq="D")
index = [pd.to_datetime(date, format='%Y-%m-%d').date() for date in index]
df.index = index
if days:
df = df.tail(days)
df = df.drop('Date_YMD', axis=1)
ax = df.plot(y=['Daily Confirmed', 'Daily Recovered', 'Daily Deceased'],
kind='line',
linewidth=2.0)
if not days or (days and days > 100):
ax.xaxis.set_major_locator(mdates.MonthLocator(interval=1))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%B %Y'))
plt.gcf().autofmt_xdate()
plt.xlabel("Month")
plt.ylabel("Number of Cases")
plt.title("COVID-19 India")
plt.savefig(buffer, format='png', pad_inches=0.1, bbox_inches='tight')
del df
def ql_errors(ql_data, out, name):
fig, ax = plt.subplots(figsize=(10, 5))
s = config['ql_params'][name]['S']
colormap = plt.cm.gist_ncar
markers = itertools.cycle((',', '+', 'x', 'o', '*', 's', 'd'))
num_plots = len(config['ql_alpha']) * len(config['ql_epsilon']) + 2
plt.gca().set_prop_cycle(
plt.cycler('color', plt.cm.jet(np.linspace(0, 3, num_plots))))
for eps, alpha in list(
ql_data[['epsilon',
'alpha']].drop_duplicates().to_records(index=False)):
#print(alpha, eps)
x = ql_data[(ql_data['epsilon'] == eps)
& (ql_data['alpha'] == alpha)][[
'error_mean', 'mean_V', 'reward', 'policy'
]]
#plt.plot(x['error_mean'].values[0], label='Error alpha='+str(alpha)+' eps='+str(eps), alpha=0.5)
plt.plot(x['error_mean'].values[0],
marker=next(markers),
markersize='5',
markevery=100,
alpha=0.5,
label='alpha=' + str(round(alpha, 3)) + ' eps=' +
str(round(eps, 3)))
#plt.plot(x['reward'].values[0], label='Reward V alpha='+str(alpha)+' eps='+str(eps), alpha=0.25)
#plt.xscale('log')
plt.grid(which='both')
plt.xlabel('iterations/' + str(config['ql_iters'][0] // 10000))
plt.ylabel('error')
plt.legend(prop={'size': 7})
plt.title('QL Error-' + name.upper())
plt.tight_layout()
plt.savefig(os.path.join(out, 'ql_error-' + name.upper() + '.png'))
plt.close()
def plotCNo(satellites):
fig, ax = plt.subplots(figsize=(18 ,15))
timestamps = list(satellites.keys())
x = range(len(timestamps))
SatIDs = satellites.values()
tmp = []
for sats in SatIDs:
for sat in sats:
if int(sat) not in tmp:
tmp.append(int(sat))
tmp.sort()
ax.set_prop_cycle(plt.cycler("color", plt.cm.hsv(np.linspace(0,1,len(tmp)))))
for count, sv in enumerate(tmp):
y = []
for sats in SatIDs:
for sat in sats:
if str(sv) in sats:
if sat == str(sv):
try:
y.append(float(sats[sat][0]))
except ValueError:
y.append(0)
break
else:
y.append(0)
break
ax.plot(x, y, label=sv)
plot_title = title + " Carrier noise ratio of Satellites in view"
ax.set(xlabel = 'Time (s)', ylabel = 'C/No (dB/Hz)', title = plot_title)
ax.set_ylim(ymin=0)
ax.set_xlim(xmin=0)
ax.legend(title="Satellite ID", loc='lower right', shadow=True, ncol=2)
fig.savefig(("../Images/" + datetitle + " Carrier noise ratio.png"), bbox_inches='tight')
plt.tight_layout()
plt.show()
print(stats.describe(mt_defaul_list))
colors = ['#e41a1c', '#377eb8', '#4daf4a', '#984ea3', '#ff7f00']
params = {'figure.facecolor': 'white',
'figure.subplot.bottom': 0.0,
'font.size': 16,
'legend.fontsize': 16,
'legend.borderpad': 0.2,
'legend.labelspacing': 0.2,
'legend.handlelength': 1.5,
'legend.handletextpad': 0.4,
'legend.borderaxespad': 0.2,
'lines.markeredgewidth': 2.0,
'lines.linewidth': 2.0,
'axes.prop_cycle': plt.cycler('color', colors)}
plt.rcParams.update(params)
fig, ax = plt.subplots(3)
ax[0].hist(predictions, 50, range=(0.0, 400.0), label='Pred', histtype='step')
ax[0].hist(target, 50, range=(0.0, 400.0), label='Truth', histtype='step')
ax[0].set_xlim([0, 400])
ax[0].set_xlabel("pT (GeV)")
ax[0].legend(loc='upper right')
ax[1].hist(result_predict, 50, range=(-100.0, 100.0), label='Pred', histtype='step')
ax[1].hist(result_default, 50, range=(-100.0, 100.0), label='Default', histtype='step')
ax[1].hist(met_truth_list, 50, range=(-100.0, 100.0), label='truth_met - met/truth_met', histtype='step')
ax[1].set_xlim([-100, 100])
ax[1].set_xlabel("Resolution (%)")
ax[1].legend(loc='upper left')
def run(job_list, output_dir):
print("Running time_series.py")
plt.style.use('fivethirtyeight')
plt.rcParams['lines.linewidth'] = 0
plt.rcParams['patch.linewidth'] = 0
plt.rcParams['axes.prop_cycle'] = plt.cycler('color', sfbi_utils.get_colors())
# disable silly warning
pd.options.mode.chained_assignment = None
df = pd.DataFrame(job_list, columns=['_id', 'city', 'user', 'submission_date', 'contract_type', 'contract_subtype'])
df2 = pd.DataFrame({'date': df.submission_date, 'type': df.contract_type})
df2_g = pd.DataFrame({'count': df2.groupby(['type', 'date']).size()}).reset_index()
df2_t = df2_g.pivot(index='date', columns='type', values='count')
df2_t.index = pd.to_datetime(df2_t.index)
df2_t2 = df2_t.resample('1M').sum()
df2_t2.columns.name = 'Type'
ax = df2_t2.plot(kind='area')
plt.title(u'Évolution des types de poste', y=1.08)
ax.set_xlabel('Date')
ax.set_ylabel("Nombre d'offres")
plt.savefig(os.path.join(output_dir, 'time_series_1.svg'), bbox_inches='tight')
plt.close()
# just CDD
df3 = pd.DataFrame({'date': df.submission_date, 'subtype': df.contract_subtype})
df3_g = pd.DataFrame({'count': df3.groupby(['subtype', 'date']).size()}).reset_index()
tmp = df3_g['subtype'].isin(['Post-doc / IR', u'CDD Ingénieur', 'CDD autre', 'ATER'])
df3_g = df3_g[tmp].reset_index()
df3_t = df3_g.pivot(index='date', columns='subtype', values='count')
df3_t.index = pd.to_datetime(df3_t.index)
df3_t2 = df3_t.resample('1M').sum()
df3_t2.columns.name = 'Type'
ax = df3_t2.plot(kind='area')
plt.title(u'Évolution des types de CDD', y=1.08)
ax.set_xlabel('Date')
ax.set_ylabel("Nombre d'offres")
plt.savefig(os.path.join(output_dir, 'time_series_2.svg'), bbox_inches='tight')
plt.close()
# just CDI
df3 = pd.DataFrame({'date': df.submission_date, 'subtype': df.contract_subtype})
df3_g = pd.DataFrame({'count': df3.groupby(['subtype', 'date']).size()}).reset_index()
tmp = df3_g['subtype'].isin(['CDI autre', 'IE', 'IR', 'PR', 'MdC', 'CR'])
df3_g = df3_g[tmp]
df3_t = df3_g.pivot(index='date', columns='subtype', values='count')
df3_t.index = pd.to_datetime(df3_t.index)
df3_t2 = df3_t.resample('1M').sum()
df3_t2.columns.name = 'Type'
ax = df3_t2.plot(kind='area')
plt.title(u'Évolution des types de CDI', y=1.08)
ax.set_xlabel('Date')
ax.set_ylabel("Nombre d'offres")
plt.savefig(os.path.join(output_dir, 'time_series_3.svg'), bbox_inches='tight')
plt.close()
df4 = pd.DataFrame({'date': pd.to_datetime(df.submission_date)})
df4_g = pd.DataFrame({'count': df4.groupby('date').size()})
df4_t = df4_g.resample('1M').sum()
plt.figure()
ax = df4_t.plot(linewidth=4)
plt.title(u"Évolution du nombre total d'offre", y=1.08)
ax.set_xlabel('Date')
ax.set_ylabel("Nombre d'offres")
plt.savefig(os.path.join(output_dir, 'time_series_4.svg'), bbox_inches='tight')
plt.close()
'''
contract type proportions
'''
# get the interesting fields and count the type for each date
df5 = pd.DataFrame({'date': df.submission_date, 'type': df.contract_type})
df5_g = pd.DataFrame({'count': df5.groupby(['type', 'date']).size()}).reset_index()
# have all the types as column and each date as one line
df5_t = df5_g.pivot(index='date', columns='type', values='count')
df5_t.index = pd.to_datetime(df5_t.index)
# make it per month
df5_t2 = df5_t.resample('1M').sum().fillna(0)
# add a total column to compute ratio
df5_t2['total'] = df5_t2.CDD + df5_t2.CDI + df5_t2.Stage + df5_t2[u'Thèse']
# replace all the values by ratios
df5_t3 = pd.DataFrame(df5_t2).apply(lambda x: (x / x['total'] * 100), axis=1).fillna(0).drop('total', 1)
df5_t3.columns.name = 'Type'
proportion_stackplot(df5_t3, output=os.path.join(output_dir, 'time_series_5.svg'), xlabel='Date',
ylabel='%des offres',
title=u"Évolution de la proportion des types de poste")
'''
contract subtype proportions for CDI
'''
# get the interesting fields only for CDI
df6 = pd.DataFrame({'date': df.submission_date, 'subtype': df.contract_subtype})[df['contract_type'] == 'CDI']
df6_g = pd.DataFrame({'count': df6.groupby(['subtype', 'date']).size()}).reset_index()
# have all the types as column and each date as one line
df6_t = df6_g.pivot(index='date', columns='subtype', values='count')
df6_t.index = pd.to_datetime(df6_t.index)
# make it per month
df6_t2 = df6_t.resample('1M').sum().fillna(0)
# add a total column to compute ratio
df6_t2['total'] = df6_t2['CDI autre'] + df6_t2.CR + df6_t2.IE + df6_t2.IR + df6_t2.MdC + df6_t2.PR
# replace all the values by ratios
df6_t3 = pd.DataFrame(df6_t2).apply(lambda x: (x / x['total'] * 100), axis=1).fillna(0).drop('total', 1)
df6_t3.columns.name = 'Type'
proportion_stackplot(df6_t3, output=os.path.join(output_dir, 'time_series_6.svg'), xlabel='Date',
ylabel='%des offres',
title=u"Évolution de la proportion des types de CDI")
'''
contract subtype proportions for CDD
'''
# get the interesting fields only for CDD
df7 = pd.DataFrame({'date': df.submission_date, 'subtype': df.contract_subtype})[df['contract_type'] == 'CDD']
# count subtype for each date
df7_g = pd.DataFrame({'count': df7.groupby(['subtype', 'date']).size()}).reset_index()
# have all the types as column and each date as one line
df7_t = df7_g.pivot(index='date', columns='subtype', values='count')
df7_t.index = pd.to_datetime(df7_t.index)
# make it per month
df7_t2 = df7_t.resample('1M').sum().fillna(0)
# add a total column to compute ratio
df7_t2['total'] = df7_t2.ATER + df7_t2[u'CDD Ingénieur'] + df7_t2['CDD autre'] + df7_t2['Post-doc / IR']
# replace all the values by ratios
df7_t3 = pd.DataFrame(df7_t2).apply(lambda x: (x / x['total'] * 100), axis=1).fillna(0).drop('total', 1)
df7_t3.columns.name = 'Type'
proportion_stackplot(df7_t3, output=os.path.join(output_dir, 'time_series_7.svg'), xlabel='Date',
ylabel='%des offres',
title=u'Évolution de la proportion des types de CDD')
# education level quantity
df8 = pd.DataFrame({'date': df.submission_date, 'type': df.contract_subtype})
df_bool_master = df8['type'].isin([u'CDD Ingénieur', 'IE'])
df_bool_phd = df8['type'].isin(['Post-doc / IR', 'PR', 'MdC', 'CR', 'IR', 'ATER'])
df_master = df8[df_bool_master]
df_master['type'] = 'Master'
df_phd = df8[df_bool_phd]
df_phd['type'] = 'PhD'
df8 = pd.DataFrame({'count': pd.concat([df_phd, df_master]).groupby(['type', 'date']).size()}).reset_index()
df8_t = df8.pivot(index='date', columns='type', values='count')
df8_t.index = pd.to_datetime(df8_t.index)
df8_t2 = df8_t.resample('1M').sum().fillna(0)
df8_t2.columns.name = 'Niveau'
ax = df8_t2.plot(kind='area')
ax.set_xlabel('Date')
ax.set_ylabel("Nombre d'offres")
plt.title(u"Nombre de CDD et CDI en fonction du niveau de diplôme", y=1.08)
plt.savefig(os.path.join(output_dir, 'time_series_8.svg'), bbox_inches='tight')
plt.close()
'''
education level proportions
'''
# add a total column to compute ratio
df9 = df8_t2
df9['total'] = df9.Master + df9.PhD
# replace all the values by ratios
df9 = pd.DataFrame(df9).apply(lambda x: (x / x['total'] * 100), axis=1).fillna(0).drop('total', 1)
proportion_stackplot(df9, output=os.path.join(output_dir, 'time_series_9.svg'), xlabel='Date', ylabel='%des offres',
title=u'Proportion des niveaux de diplôme requis pour les CDD et CDI')
'''
average year series
'''
df10 = pd.DataFrame({'date': df.submission_date,
'type': df.contract_type}) # .reset_index().groupby(['date'])['index'].count().reset_index(name='count')
df10 = df10.set_index(pd.DatetimeIndex(df10['date']))
df10['month'] = df10.index.month
df10 = (df10['2013':'2016'].groupby(['month', 'type'])['month'].count() / 4).reset_index(name='count')
df10_t = df10.pivot(index='month', columns='type', values='count').fillna(0)
df10_t.columns.name = 'Type'
ax = df10_t.plot(linewidth=4)
ax.set_xlabel('Mois')
ax.set_ylabel("Nombre d'offres")
plt.title(u'Moyennes mensuelles 2013-2016', y=1.08)
ax.legend(loc='center right', bbox_to_anchor=(1.25, 0.5))
plt.savefig(os.path.join(output_dir, 'time_series_10.svg'), bbox_inches='tight')
plt.close()
CDD_subtypes = ['Post-doc / IR', u'CDD Ingénieur', 'CDD autre', 'ATER']
df11 = pd.DataFrame({'date': df.submission_date, 'type': df.contract_subtype})
tmp = df11['type'].isin(CDD_subtypes)
df11 = df11[tmp]
df11 = df11.set_index(pd.DatetimeIndex(df11['date']))
df11['month'] = df11.index.month
df11 = (df11['2013':'2016'].groupby(['month', 'type'])['month'].count() / 4).reset_index(name='count')
df11_t = df11.pivot(index='month', columns='type', values='count').fillna(0)
df11_t.columns.name = 'Type'
ax = df11_t.plot(linewidth=4)
ax.set_xlabel('Mois')
ax.set_ylabel("Nombre d'offres")
plt.title(u'Moyennes mensuelles 2013-2016 pour les CDD', y=1.08)
ax.legend(loc='center right', bbox_to_anchor=(1.4, 0.5))
plt.savefig(os.path.join(output_dir, 'time_series_11.svg'), bbox_inches='tight')
plt.close()
CDI_subtypes = ['CDI autre', 'IE', 'IR', 'PR', 'MdC', 'CR']
df12 = pd.DataFrame({'date': df.submission_date, 'type': df.contract_subtype})
tmp = df12['type'].isin(CDI_subtypes)
df12 = df12[tmp]
df12 = df12.set_index(pd.DatetimeIndex(df12['date']))
df12['month'] = df12.index.month
df12 = (df12['2013':'2016'].groupby(['month', 'type'])['month'].count() / 4).reset_index(name='count')
df12_t = df12.pivot(index='month', columns='type', values='count').fillna(0)
df12_t.columns.name = 'Type'
ax = df12_t.plot(linewidth=4)
plt.legend(loc=7)
ax.set_xlabel('Mois')
ax.set_ylabel("Nombre d'offres")
plt.title(u'Moyennes mensuelles 2013-2016 pour les CDI', y=1.08)
ax.legend(loc='center right', bbox_to_anchor=(1.3, 0.5))
plt.savefig(os.path.join(output_dir, 'time_series_12.svg'), bbox_inches='tight')
plt.close()
def plotSSCurve(sensitivity, specificity, people=[], title="", show_legend=False, textcolor='#4D5B66', plot_results=True):
"""SS curve for Publications.
Args:
sensitivity (numpy.array): length-N array of sensitivity points.
specificity (numpy.array): length-N array of specificity points.
people (list): List of lists. Each list entry is two float values:
[persons_sensitivity, persons_specificity]
Returns
the handle to the figure, for further plotting.
"""
area = auc(sensitivity, specificity)
print 'The AUC is %s' % area
# textcolor = 'black'
# textcolor = 'darkgrey'
# textcolor = '#4D5B66'
textsize = 24
rcParams = {
'axes.grid' : False,
'font.family' : 'sans-serif',
'text.color' : textcolor,
'axes.labelcolor' : textcolor,
'axes.labelsize' : textsize,
'axes.titlesize' : textsize,
'axes.facecolor' : 'white',
'axes.linewidth' : 3,
'axes.prop_cycle' : plt.cycler('color', ['blue', 'black', '#5BC0DE', 'blue']),
'figure.figsize' : (8,8),
'xtick.color' : textcolor,
'xtick.labelsize' : 20,
'xtick.major.pad' : 15,
'ytick.color' : textcolor,
'ytick.labelsize' : 20,
'ytick.major.pad' : 15,
'legend.fontsize' : 20,
}
with plt.rc_context(rcParams):
fig = plt.figure()
ax = fig.add_subplot(111)
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
plt.plot(sensitivity, specificity, label='Algorithm: AUC=%0.2f' % area, linewidth=4)
xlabel = 'Sensitivity'
ylabel = 'Specificity'
plt.xlabel(xlabel)
plt.ylabel(ylabel)
plt.ylim([0.0, 1.05])
plt.xlim([0.0, 1.05])
plt.yticks([0.0, 1.0], ["0", "1"])
plt.xticks([0.0, 1.0], ["0", "1"])
plt.axhline(y=1.0, xmin=0, xmax=0.95, color='k', zorder=-34, ls='dashed')
plt.axhline(y=0.9, xmin=0, xmax=0.95, color='k', zorder=-34, ls='dashed')
plt.axhline(y=0.8, xmin=0, xmax=0.95, color='k', zorder=-34, ls='dashed')
plt.axvline(x=1.0, ymin=0, ymax=0.95, color='k', zorder=-34, ls='dashed')
plt.axvline(x=0.9, ymin=0, ymax=0.95, color='k', zorder=-34, ls='dashed')
plt.axvline(x=0.8, ymin=0, ymax=0.95, color='k', zorder=-34, ls='dashed')
area = "%0.2f" % area
title = title + '\n'
plt.title(title)
for i, person in enumerate(people):
if i == 0:
plt.plot(person[0], person[1], 'o', color='red', zorder=-32, markersize=10, label='Dermatologists (%d)' % len(people))
else:
plt.plot(person[0], person[1], 'o', color='red', zorder=-32, markersize=10)
if len(people) > 0:
avg_sensitivity = np.mean(np.array(people)[:,0])
avg_specificity = np.mean(np.array(people)[:,1])
std_sensitivity = np.std(np.array(people)[:,0])
std_specificity = np.std(np.array(people)[:,1])
print 'Average sensitivity=%0.2f +- %0.2f' % (avg_sensitivity, std_sensitivity)
print 'Average specificity=%0.2f +- %0.2f' % (avg_specificity, std_specificity)
plt.plot(avg_sensitivity, avg_specificity, 'D', color='green', markersize=10, label='Average Dermatologist')
plt.errorbar(avg_sensitivity, avg_specificity, xerr=std_sensitivity, yerr=std_specificity,
color='green', markersize=10, elinewidth=3)
if show_legend:
plt.legend(loc='lower left', numpoints= 1)
if plot_results:
plt.show()
return fig, ax
def plot_species(df, species_list=None, save_name='test', out_dir=None,
title=None, plot_type='plotly', image_format='pdf',
close_plots=False):
"""
Parameters
----------
df: pandas.DataFrame
magine formatted dataframe
species_list: list
List of genes to be plotter
save_name: str
Filename to be saved as
out_dir: str
Path for output to be saved
title: str
Title of plot, useful when list of genes corresponds to a GO term
plot_type : str
Use plotly to generate html output or matplotlib to generate pdf
image_format : str
pdf or png, only used if plot_type="matplotlib"
close_plots : bool
Close plot after making, use when creating lots of plots in parallel.
Returns
-------
"""
ldf = df.copy()
if out_dir is not None:
if not os.path.exists(out_dir):
os.mkdir(out_dir)
# gather x axis points
x_points = sorted(ldf[sample_id].unique())
if len(x_points) == 0:
return
if isinstance(x_points[0], np.float):
x_point_dict = {i: x_points[n] for n, i
in enumerate(x_points)}
else:
x_point_dict = {i: n for n, i
in enumerate(x_points)}
if species_list is not None:
ldf = ldf.loc[ldf[identifier].isin(species_list)].copy()
ldf = log2_normalize_df(ldf, column=fold_change)
n_plots = len(ldf[identifier].unique())
num_colors = len(ldf[label_col].unique())
color_list = sns.color_palette("tab20", num_colors)
if plot_type == 'matplotlib':
fig = plt.figure()
ax = fig.add_subplot(111)
ax.set_prop_cycle(plt.cycler('color', color_list))
colors = enumerate(color_list)
plotly = []
names_list = []
total_counter = 0
for name, j in ldf.groupby(identifier):
index_counter = 0
for n, m in j.groupby(label_col):
x = np.array(m[sample_id])
if len(x) < 1:
continue
y = np.array(m['fold_change'])
sig_flag = np.array(m[flag])
index = np.argsort(x)
x = x[index]
y = y[index]
s_flag = sig_flag[index]
# x values with scaled values (only changes things if non-float
# values are used for sample_id
x_index = np.array([x_point_dict[ind] for ind in x])
index_counter += 1
total_counter += 1
# create matplotlib plot
if plot_type == 'matplotlib':
label = "\n".join(wrap(n, 40))
p = ax.plot(x_index, y, '.-', label=label)
if len(s_flag) != 0:
color = p[0].get_color()
ax.plot(x_index[s_flag], y[s_flag], '^', color=color)
# create plotly plot
elif plot_type == 'plotly':
c = next(colors)[1]
plotly.append(_ploty_graph(x_index, y, n, n, c))
if len(s_flag) != 0:
index_counter += 1
total_counter += 1
plotly.append(_ploty_graph(x_index[s_flag], y[s_flag],
n, n, c, marker='x-open-dot'))
names_list.append([name, index_counter])
if plot_type == 'matplotlib':
lgd = _format_mpl(ax, x_point_dict, x_points)
if save_name is not None:
tmp_savename = "{}.{}".format(save_name, image_format)
if out_dir is not None:
tmp_savename = os.path.join(out_dir, tmp_savename)
plt.savefig(tmp_savename, bbox_extra_artists=(lgd,),
bbox_inches='tight')
if close_plots:
plt.close(fig)
else:
return fig
elif plot_type == 'plotly':
fig = _create_plotly(total_counter, n_plots, names_list, x_point_dict,
title, x_points, plotly)
if save_name:
_save_ploty_output(fig, out_dir, save_name)
else:
init_notebook_mode(connected=True)
iplot(fig)
def run(job_list, output_dir):
print("Running summary.py")
contract_subtype_level = {'Post-doc / IR': 'PhD',
'PR': 'PhD',
'MdC': 'PhD',
'CR': 'PhD',
'IR': 'PhD',
'ATER': 'PhD',
u'CDD Ingénieur': 'Master',
'IE': 'Master'}
colors = sfbi_utils.get_colors()
plt.style.use('fivethirtyeight')
plt.rcParams['axes.prop_cycle'] = plt.cycler('color', sfbi_utils.get_colors())
# for the pie charts
plt.rcParams['patch.linewidth'] = 1
plt.rcParams['patch.edgecolor'] = 'white'
df = pd.DataFrame(job_list, columns=['_id', 'contract_type', 'contract_subtype', 'city', 'submission_date'])
summary_file = open(os.path.join(output_dir, 'summary.txt'), 'w')
submission_date_series = pd.Series(df.submission_date)
summary_file.write("From " + str(submission_date_series.min()) + " to " + str(submission_date_series.max()) + '\n')
summary_file.write("Total jobs: " + str(len(df.index)) + "\n")
summary_file.write(pd.Series(df.contract_type).value_counts().to_string(header=False) + "\n")
# general types ratios pie chart
df_contract_type_count = pd.Series(df.contract_type).value_counts(sort=True)
fig, ax = minimal_hbar(df_contract_type_count)
ax.set_title(u'Types de postes')
fig.savefig(os.path.join(output_dir, 'summary_6.svg'), bbox_inches='tight')
plt.close(fig)
plt.figure()
ax = df_contract_type_count.plot(kind='pie', startangle=90, autopct='%1.1f%%', colors=colors)
ax.set_xlabel('')
ax.set_ylabel('')
plt.axis('equal')
plt.title('Types de poste', y=1.08)
plt.savefig(os.path.join(output_dir, 'summary_1.svg'), bbox_inches='tight')
plt.close()
# CDI subtypes ratios pie chart
df_contract_subtype_CDI_count = pd.Series(df[df.contract_type == 'CDI'].contract_subtype).value_counts(sort=True)
fig, ax = minimal_hbar(df_contract_subtype_CDI_count)
ax.set_title(u'Types de CDI')
fig.savefig(os.path.join(output_dir, 'summary_7.svg'), bbox_inches='tight')
plt.close(fig)
plt.figure()
ax = df_contract_subtype_CDI_count.plot(kind='pie', startangle=90, autopct='%1.1f%%', colors=colors)
ax.set_xlabel('')
ax.set_ylabel('')
plt.axis('equal')
plt.title('Types de CDI', y=1.08)
plt.savefig(os.path.join(output_dir, 'summary_2.svg'), bbox_inches='tight')
plt.close()
# CDD subtypes ratios pie chart
df_contract_subtype_CDD_count = pd.Series(df[df.contract_type == 'CDD'].contract_subtype).value_counts(sort=True)
fig, ax = minimal_hbar(df_contract_subtype_CDD_count)
ax.set_title(u'Types de CDD')
fig.savefig(os.path.join(output_dir, 'summary_8.svg'), bbox_inches='tight')
plt.close(fig)
plt.figure()
ax = df_contract_subtype_CDD_count.plot(kind='pie', startangle=90, autopct='%1.1f%%', colors=colors)
ax.set_xlabel('')
ax.set_ylabel('')
plt.axis('equal')
plt.title('Types de CDD', y=1.08)
plt.savefig(os.path.join(output_dir, 'summary_3.svg'), bbox_inches='tight')
plt.close()
# best cities
df_city = pd.Series(df.city).value_counts()
# df_city = df_city[df_city >= 10]
sum_small_cities = df_city.iloc[16:].sum()
df_city = df_city.iloc[:15]
df_city['Autres'] = sum_small_cities
# plt.figure()
city_names = df_city.index
labels = ['{0} - {1:1.2f} %'.format(i, j) for i, j in zip(city_names, 100. * df_city / df_city.sum())]
# [' '] * len(labels) -> empty labels for each pie, to avoid matplotlib warning
axes = pd.DataFrame(df_city).plot(kind='pie', startangle=90, labels=[' '] * len(labels), colors=colors,
subplots=True)
for ax in axes:
ax.legend(labels, loc=1, bbox_to_anchor=(1.5, 1.1))
ax.set_xlabel('')
ax.set_ylabel('')
plt.axis('equal')
plt.title("Parts des 15 villes ayant le plus d'offres d'emploi", y=1.08)
plt.savefig(os.path.join(output_dir, 'summary_4.svg'), bbox_inches='tight')
plt.close()
# education level ratios pie chart
df_study_level = pd.Series(df.contract_subtype).value_counts(sort=True)
level_dict = {'PhD': 0, 'Master': 0}
for subtype in df_study_level.index:
if subtype in contract_subtype_level:
level = contract_subtype_level[subtype]
level_dict[level] += df_study_level[subtype]
plt.figure()
ax = pd.Series(level_dict).plot(kind='pie', startangle=90, autopct='%1.1f%%', colors=colors)
ax.set_xlabel('')
ax.set_ylabel('')
plt.axis('equal')
plt.title(u'Niveau de diplôme requis pour les CDD et CDI', y=1.08)
plt.savefig(os.path.join(output_dir, 'summary_5.svg'), bbox_inches='tight')
plt.close()
# per city contract types
city_gb = df.groupby('city')
df_type_city = city_gb.contract_type.value_counts(sort=True, ascending=True).unstack(level=-1).fillna(0)
df_type_city['Total'] = df_type_city.sum(axis='columns')
df_perc_city = df_type_city.div(df_type_city['Total'], axis='index')
df_perc_city['Total'] = df_type_city['Total']
df_perc_city = df_perc_city.sort_values(by='Total')
fig, ax = plt.subplots()
ax = df_perc_city.loc[:, ['CDD', 'CDI', 'Stage', u'Thèse']].iloc[-10:].plot(ax=ax, kind='barh', stacked=True,
color=colors)
ax.set_xlim([0, 1])
ax.set_ylabel('')
i = 0
for total in df_perc_city['Total'].iloc[-10:]:
ax.text(1.02, i, int(total), va='center')
i += 1
plt.title(u"Répartition des types d'offres\ndans les 10 villes ayant le plus d'offres", y=1.08)
ax.legend(loc='lower center', bbox_to_anchor=(0.5, -0.2), ncol=4)
fig.savefig(os.path.join(output_dir, 'summary_9.svg'), bbox_inches='tight')
plt.close(fig)
# per city education level
def get_jobs_educ_level(series):
level_dict = {'PhD': 0, 'Master': 0}
for e in series.index:
if e in contract_subtype_level:
level_dict[contract_subtype_level[e]] += series[e]
return pd.Series(level_dict)
def count_master(series):
return get_jobs_educ_level(series.value_counts())['Master']
def count_phd(series):
return get_jobs_educ_level(series.value_counts())['PhD']
df_level_city = city_gb.contract_subtype.aggregate({'Master': count_master, 'PhD': count_phd})
df_level_city['Total'] = df_level_city['Master'] + df_level_city['PhD']
df_level_perc = df_level_city.div(df_level_city['Total'], axis='index')
df_level_perc['Total'] = df_level_city['Total']
df_level_perc = df_level_perc.sort_values(by='Total')
fig, ax = plt.subplots()
ax = df_level_perc.loc[:, ['Master', 'PhD']].iloc[-10:].plot(ax=ax, kind='barh', stacked=True, color=colors)
ax.set_xlim([0, 1])
ax.set_ylabel('')
i = 0
for total in df_level_perc['Total'].iloc[-10:]:
ax.text(1.02, i, int(total), va='center')
i += 1
plt.title(u"Proportion des niveaux de diplôme requis\ndans les 10 villes ayant le plus d'offres", y=1.08)
ax.legend(loc='lower center', bbox_to_anchor=(0.5, -0.2), ncol=4)
fig.savefig(os.path.join(output_dir, 'summary_10.svg'), bbox_inches='tight')
plt.close(fig)
############################################################################
# Calculation with Berreman4x4
data = Berreman4x4.DataList([s.evaluate(kx,k0) for kx in Kx])
R_p = data.get('R_pp')
R_s = data.get('R_ss')
T_p = data.get('T_pp')
T_s = data.get('T_ss')
t2_p = abs(data.get('t_pp'))**2 # Before power correction
t2_s = abs(data.get('t_ss'))**2
############################################################################
# Plotting
fig = pyplot.figure(figsize=(12., 6.))
pyplot.rcParams['axes.prop_cycle'] = pyplot.cycler('color', 'bgrcbg')
ax = fig.add_axes([0.1, 0.1, 0.7, 0.8])
y = numpy.vstack((R_s,R_p,t2_s,t2_p,T_s,T_p)).T
legend1 = ("R_s","R_p","t2_s","t2_p","T_s","T_p")
lines1 = ax.plot(Kx, y)
y_th = numpy.vstack((R_th_s, R_th_p, t2_th_s, t2_th_p,
T_th_s, T_th_p)).T
legend2 = ("R_th_s", "R_th_p", "t2_th_s", "t2_th_p",
"T_th_s", "T_th_p")
lines2 = ax.plot(Kx, y_th, 'x')
ax.legend(lines1 + lines2, legend1 + legend2,
loc='upper left', bbox_to_anchor=(1.05, 1), borderaxespad=0.)
lumi_par_tot2+=element
lumi_tot2 = lumi_par_tot2/13.
#print lumi_tot
#print lumi_tot2
storage.append(lumi_tot/lumi_tot2)
channel_excl.append(storage)
#print channel_excl
print "Done with the file"
print "Now ready to plot"
#this will set different colors in the same plot
num_plots = 14
colormap = plt.cm.gist_ncar
plt.gca().set_prop_cycle(plt.cycler('color', plt.cm.gist_ncar(np.linspace(0, 0.9, num_plots))))
x=range(total_nb4)
labels=[["ch1", "ch2", "ch3", "ch5", "ch6"],[ "ch7", "ch8", "ch9", "ch10", "ch11"],[ "ch12", "ch13", "ch14", "ch15"]]
lgd=0
for alpha in range(14):
y=[]
for beta in range(total_nb4):
y.append(channel_excl[beta][alpha])
plt.plot(x,y)
if alpha==4 or alpha==9 or alpha==13:
plt.legend(labels[lgd], loc = "best", fancybox=True, shadow=True)
plt.show()
lgd+=1
self.blue = '#3498db'
self.red = '#e74c3c'
self.green = '#2ecc71'
self.orange = '#e67e22'
self.yellow = '#f1c40f'
self.purple = '#9b59b6'
self.gray = '#95a5a6'
self.teal = '#1abc9c'
self.bluegray = '#34495e'
self.dark_blue = '#2980b9'
self.dark_red = '#c0392b'
self.dark_green = '#27ae60'
self.dark_orange = '#d35400'
self.dark_yellow = '#f39c12'
self.dark_purple = '#8e44ad'
self.dark_gray = '#7f8c8d'
self.dark_teal = '#16a085'
self.dark_bluegray = '#2c3e50'
colors = _Colors()
color_sequence = ['blue', 'red', 'green', 'orange', 'purple', 'yellow', 'gray',
'teal', 'bluegray']
color_cycle = plt.cycler('color', [colors.__dict__[s] for s in color_sequence])
color_cycle += plt.cycler('mec', [colors.__dict__['dark_{}'.format(s)]
for s in color_sequence])
style = {'axes.prop_cycle': color_cycle}
def cf(date_dir, atmfile, filters, plot=True):
"""
Call above functions to calculate cf and plot them
"""
# Read atmfile
molecules, p, T, abundances = mat.readatm(date_dir + atmfile)
nlayers = len(p)
# Read tau.dat
foo = date_dir + 'tau.dat'
tau, wns = readTauDat(foo, nlayers)
# Calculate BB, reverse the order of p and T
p = p[::-1]
T = T[::-1]
BB = Planck(T, wns)
# Call cf_eq() to calculate cf
cf = cf_eq(BB, p, tau, nlayers, wns)
# Call filter_cf() to calculate cf
filt_cf, filt_cf_norm = filter_cf(filters, nlayers, wns, cf, normalize=True)
if plot:
print(" Plotting contribution functions.\n")
# Not normalized cf
plt.figure(4)
plt.clf()
gs = gridspec.GridSpec(1, 2, width_ratios=[5, 1])
ax0 = plt.subplot(gs[0])
colormap = plt.cm.rainbow(np.linspace(0, 1, len(filters)))
ax0.set_prop_cycle(plt.cycler('color', colormap))
for i in np.arange(len(filt_cf)):
(head, tail) = os.path.split(filters[i])
lbl = tail[:-4]
ax0.semilogy(filt_cf[i], p, '-', linewidth = 1, label=lbl)
lgd = ax0.legend(loc='center left', bbox_to_anchor=(1.0, 0.5),
ncol=len(filt_cf)//30 + 1, prop={'size':8})
ax0.set_ylim(max(p), min(p))
ax0.ticklabel_format(style='sci', axis='x', scilimits=(0,0))
ax0.set_xlabel('Contribution Functions', fontsize=14)
ax0.set_ylabel('Pressure (bar)' , fontsize=14)
plt.savefig(date_dir + 'ContrFuncs.png', bbox_extra_artists=(lgd,),
bbox_inches='tight')
# Normalized cf
plt.figure(5)
plt.clf()
gs = gridspec.GridSpec(1, 2, width_ratios=[5, 1])
ax0 = plt.subplot(gs[0])
colormap = plt.cm.rainbow(np.linspace(0, 1, len(filters)))
ax0.set_prop_cycle(plt.cycler('color', colormap))
for i in np.arange(len(filt_cf_norm)):
(head, tail) = os.path.split(filters[i])
lbl = tail[:-4]
ax0.semilogy(filt_cf_norm[i], p, '--', linewidth = 1, label=lbl)
lgd = ax0.legend(loc='center left', bbox_to_anchor=(1,0.5),
ncol=len(filt_cf)//30 + 1, prop={'size':8})
ax0.set_ylim(max(p), min(p))
ax0.set_xlim(0, 1.0)
ax0.ticklabel_format(style='sci', axis='x', scilimits=(0,0))
ax0.set_xlabel('Normalized Contribution Functions', fontsize=14)
ax0.set_ylabel('Pressure (bar)' , fontsize=14)
plt.savefig(date_dir + 'NormContrFuncs.png', bbox_extra_artists=(lgd,),
bbox_inches='tight')
return filt_cf[:,::-1], filt_cf_norm[:,::-1]
Ef = 0.63 # DOS at Fermi level in states/eV
a = 5.6*5.292*(10**(-11))
epsilon = 8.854*10**(-12) # Permittivity of free space
munot = 4*np.pi*10**(-7) # vacuum permeability
hbar = 6.58*(10**(-13)) # reduced Plank's constant
T = 200 #K
# Boltzmann's constant
k = 8.6173303*10**(-5) # in eV.K-1
mu = 0.0 # chemical potential
N0 = Ef/(e*0.5*a**3)
conduct_coeff = N0*(0.33*v**2)*(e**2)*hbar
w_p = np.sqrt(N0*(0.33*v**2)*(e**2)/epsilon)*(hbar/1000) # plasma frequency in eV
#######################################################################################################################
colormap = plt.cm.plasma
plt.gca().set_prop_cycle(plt.cycler('color', colormap(np.linspace(0, 0.85, num_plots))))
print ("Conductance coefficient:", conduct_coeff, "Plasma Frequency:", w_p)
# Define your own input arrays - can be different (though still within the bounds) than those in the files
# This gives us the flexibility to check for intergation convergence
# We accomplish this by interpolating, soon to follow
#Note: As mentioned above, choose a spacing and endpoints such that w'+w also maps to the w'_array for every w.
# Otherwise one will have to interpolate the self-energy on the w'+w grid for every value of w. that will be unwise.
array_omegap = np.arange(-5, 3, delta_omegap) # The energy grid for the self-energy.
array_omega_a = np.arange(0.002, 0.02, delta_omega) # The energy grid for the incoming infrared photon
array_omega_b = np.arange(0.02, 0.80, delta_omega) # The energy grid for the incoming infrared photon
array_omega = np.concatenate((array_omega_a, array_omega_b))
array_M = interpolate(path_to_realpart, array_omegap)/1000.0
array_Gamma = interpolate(path_to_linewidth, array_omegap)/1000.0
def callTransit(atmfile, tepfile, MCfile, stepsize, molfit, solution, p0,
tconfig, date_dir, burnin, abun_file, PTtype, PTfunc,
filters, ctf=None):
"""
Call Transit to produce best-fit outputs.
Plot MCMC posterior PT plot.
Parameters:
-----------
atmfile: String
Atmospheric file.
tepfile: String
Transiting extra-solar planet file.
MCfile: String
File with MCMC log and best-fitting results.
stepsize: 1D float ndarray
Specified step sizes for each parameter.
molfit: 1D String ndarray
List of molecule names to modify their abundances.
solution: String
Flag to indicate transit or eclipse geometry
p0: Float
Atmosphere's 'surface' pressure level.
tconfig: String
Transit configuration file.
date_dir: String
Directory where to store results.
burnin: Integer
abun_file: String
Elemental abundances file.
PTtype: String
Pressure-temperature profile type ('line' for Line et al. 2013,
'madhu_noinv' or 'madhu_inv' for Madhusudhan & Seager 2009,
'iso' for isothermal)
PTfunc: pointer to function
Determines the method of evaluating the PT profile's temperature
filters: list, strings.
Filter files associated with the eclipse/transit depths
ctf: 2D array.
Contribution or transmittance functions corresponding to `filters`
"""
# make sure burnin is an integer
burnin = int(burnin)
# read atmfile
molecules, pressure, temp, abundances = mat.readatm(atmfile)
# get surface gravity
grav, Rp = mat.get_g(tepfile)
# get star data if needed
if PTtype == 'line':
R_star, T_star, sma, gstar = get_starData(tepfile)
# FINDME: Hardcoded value:
T_int = 100 # K
PTargs = [R_star, T_star, T_int, sma, grav*1e2]
else:
PTargs = None # For non-Line profiles
# Get best parameters
bestP, uncer = read_MCMC_out(MCfile)
allParams = bestP
# get PTparams and abundances factors
nparams = len(allParams)
nmol = len(molfit)
nradfit = int(solution == 'transit')
nPTparams = nparams - nmol - nradfit
PTparams = allParams[:nPTparams]
# call PT profile generator to calculate temperature
best_T = pt.PT_generator(pressure, PTparams, PTfunc, PTargs)
# Plot best PT profile
plt.figure(1)
plt.clf()
plt.semilogy(best_T, pressure, '-', color = 'r')
plt.xlim(0.9*min(best_T), 1.1*max(best_T))
plt.ylim(max(pressure), min(pressure))
plt.title('Best PT', fontsize=14)
plt.xlabel('T (K)' , fontsize=14)
plt.ylabel('logP (bar)', fontsize=14)
# Save plot to current directory
plt.savefig(date_dir + 'Best_PT.png')
# Update R0, if needed:
if nradfit:
Rp = allParams[nPTparams]
# write best-fit atmospheric file
write_atmfile(atmfile, abun_file, molfit, best_T,
allParams[nPTparams+nradfit:], date_dir, p0, Rp, grav)
# write new bestFit Transit config
if solution == 'transit':
bestFit_tconfig(tconfig, date_dir, allParams[nPTparams])
else:
bestFit_tconfig(tconfig, date_dir)
# Call Transit with the best-fit tconfig
Transitdir = os.path.dirname(os.path.realpath(__file__)) + \
"/../modules/transit/"
bf_tconfig = date_dir + 'bestFit_tconfig.cfg'
Tcall = Transitdir + "/transit/transit"
subprocess.call(["{:s} -c {:s}".format(Tcall, bf_tconfig)],
shell=True, cwd=date_dir)
# ========== plot MCMC PT profiles ==========
# get MCMC data:
MCMCdata = date_dir + "/output.npy"
data = np.load(MCMCdata)
nchains, npars, niter = np.shape(data)
# stuck chains:
data_stack = data[0,:,burnin:]
for c in np.arange(1, nchains):
data_stack = np.hstack((data_stack, data[c, :, burnin:]))
# create array of PT profiles
PTprofiles = np.zeros((np.shape(data_stack)[1], len(pressure)))
# current PT parameters for each chain, iteration
curr_PTparams = PTparams
# fill-in PT profiles array
if ctf is None:
print(" Plotting MCMC PT profile figure.")
for k in np.arange(0, np.shape(data_stack)[1]):
j = 0
for i in np.arange(len(PTparams)):
if stepsize[i] != 0.0:
curr_PTparams[i] = data_stack[j,k]
j +=1
else:
pass
PTprofiles[k] = pt.PT_generator(pressure, curr_PTparams,
PTfunc, PTargs)
# get percentiles (for 1,2-sigma boundaries):
low1 = np.percentile(PTprofiles, 16.0, axis=0)
hi1 = np.percentile(PTprofiles, 84.0, axis=0)
low2 = np.percentile(PTprofiles, 2.5, axis=0)
hi2 = np.percentile(PTprofiles, 97.5, axis=0)
median = np.median( PTprofiles, axis=0)
# plot figure
plt.figure(2, dpi=300)
plt.clf()
if ctf is not None:
plt.subplots_adjust(wspace=0.15)
ax1=plt.subplot(121)
else:
ax1=plt.subplot(111)
ax1.fill_betweenx(pressure, low2, hi2, facecolor="#62B1FF",
edgecolor="0.5")
ax1.fill_betweenx(pressure, low1, hi1, facecolor="#1873CC",
edgecolor="#1873CC")
plt.semilogy(median, pressure, "-", lw=2, label='Median',color="k")
plt.semilogy(best_T, pressure, "-", lw=2, label="Best fit", color="r")
plt.ylim(pressure[0], pressure[-1])
plt.legend(loc="best")
plt.xlabel("Temperature (K)", size=15)
plt.ylabel("Pressure (bar)", size=15)
if ctf is not None:
# Add contribution or transmittance functions
ax2=plt.subplot(122, sharey=ax1)
colormap = plt.cm.rainbow(np.linspace(0, 1, len(filters)))
ax2.set_prop_cycle(plt.cycler('color', colormap))
# Plot with filter labels
for i in np.arange(len(filters)):
(head, tail) = os.path.split(filters[i])
lbl = tail[:-4]
ax2.semilogy(ctf[i], pressure, '--', linewidth = 1, label=lbl)
# Hide y axis tick labels
plt.setp(ax2.get_yticklabels(), visible=False)
# Place legend off figure in case there are many filters
lgd = ax2.legend(loc='center left', bbox_to_anchor=(1,0.5),
ncol=len(filters)//30 + 1, prop={'size':8})
if solution == 'eclipse':
ax2.set_xlabel('Normalized Contribution\nFunctions', fontsize=15)
else:
ax2.set_xlabel('Transmittance', fontsize=15)
# save figure
if ctf is not None:
savefile = date_dir + "MCMC_PTprofiles_cf.png"
plt.savefig(savefile, bbox_extra_artists=(lgd,), bbox_inches='tight')
else:
savefile = date_dir + "MCMC_PTprofiles.png"
plt.savefig(savefile)