def plot_sentiment_by_country(data_mails, opt, nb_country=20):
'''
This function plots sentiment of countries (countries are filtered by sentiment).
This function uses 4 colors to evaluate the occurrences of given countries.
Abscissa: countries / Ordinate: sentiment / Color: occurrences
Parameters
- opt : allow to select the option of the graph
- 'only_good' = keep only good feeling
- 'only_bad' = keep only bad feeling
- <empty> = concatenate good and bad feeling in same plot
- data_mails : DataFrame containing the data (emails)
- nb_country : selection of the most representative countries (by default, 20)
'''
# Selection of data to be used for the plot
if opt == 'only_good':
data_plot = data_mails.nlargest(nb_country, 'Sentiment')
title = 'Hilary\'s opinion for the ' + str(
nb_country) + ' most-preferred countries'
elif opt == 'only_bad':
data_plot = data_mails.nsmallest(nb_country, 'Sentiment')
title = 'Hilary\'s opinion for the ' + str(
nb_country) + ' less-preferred countries'
else:
most_liked = data_mails.nlargest(nb_country, 'Sentiment')
worst_liked = data_mails.nsmallest(nb_country, 'Sentiment')
data_plot = pd.concat([most_liked, worst_liked])
title = 'Hilary\'s opinion for the ' + str(
nb_country) + ' most- and less-preferred countries'
data_plot.sort_values(by='Sentiment', ascending=False, inplace=True)
# Definition of the gradation of color
y = np.array(np.log(data_plot['Occurrences']))
colors = cm.BrBG(y / float(max(y)))
plot = plt.scatter(y, y, c=y, cmap='BrBG')
plt.clf()
clb = plt.colorbar(plot)
clb.ax.set_title('Occurences (log)')
# Creation of the plot
sentiment_data_plot = sns.barplot(x=data_plot.index,
y='Sentiment',
data=data_plot,
palette=colors)
# Display of a line to separate the graph (static line, for delimitation of 20 most- and less-preferred countries)
define_plot_title_and_labels(sentiment_data_plot, title=title)
if opt == None and nb_country == 20:
sentiment_data_plot.axvline(nb_country - 0.5)
sns.plt.show()
def getECColour(_log, EC):
_log.info("getECColour(_log, EC) EC = " + str(EC))
#normalize item number values to colormap
norm = matplotlib.colors.Normalize(vmin=0, vmax=2000)
#colormap possible values = viridis, jet, spectral,gist_rainbow, BrBG
rgba_color = cm.BrBG(norm(int(EC)), bytes=True)
_log.info("rgba_color =" + str(rgba_color))
c = Color(rgb=(rgba_color[0] / 255, rgba_color[1] / 255,
rgba_color[2] / 255))
_log.info("getECColour =" + str(c))
return c
def show_graph(g, pos, partition_dic, colorful, str_type):
n = g.number_of_nodes()
plt.figure(figsize=(30, 30))
# Read color for points
color_max = len(partition_dic.keys())
# Different color for different clusters/partition
color_index = 0
color_general_index = 0
# Blue edges
nx.draw_networkx_edges(g, pos, edge_color='b', alpha=0.3)
for represent in partition_dic.keys():
# if it's a little cluster/partition -> more transparency
if len(partition_dic[represent]) > 10:
alpha = 0.8
else:
alpha = 0.1
# If there are no more than 5 partitions with size bigger than 1 then print with colorful layout
if colorful and len(partition_dic[represent]) > 1:
color = color_general[color_general_index]
color_general_index += 1
alpha = 0.8
else:
color = cm.BrBG(color_index / color_max)
# If type = dbscan -> the cluster with 1 point will not be drawn
if str_type != "dbscan" or len(partition_dic[represent]) > 2:
for point in partition_dic[represent]:
try:
nx.draw_networkx_nodes(g,
pos,
nodelist=[point],
node_color=[color],
node_size=100,
alpha=alpha)
except nx.exception.NetworkXError:
continue
color_index += 1
# Figure settings
plt.axis('off')
plt.show(block=False)
plt.savefig(sys.argv[1].replace("edges.txt", str_type + ".png"))
return 0
def draw(nb_points, xs, ys, froms, tos, partition_dic, fig):
# Check if point is not noise (cluster size < 2)
points = []
for represent in partition_dic.keys():
for point in partition_dic[represent]:
points.append(point)
if len(points) == 0:
print("No cluster detected!")
return 1
# Print edges
for edge in range(len(froms)):
from_loc = froms[edge]
to_loc = tos[edge]
if from_loc in points and to_loc in points:
fig.arrow(xs[from_loc],
ys[from_loc],
xs[to_loc] - xs[from_loc],
ys[to_loc] - ys[from_loc],
color='m',
alpha=0.1)
# Print points
color_max = len(partition_dic.keys())
color_index = 0
for represent in partition_dic.keys():
r, g, b, a = cm.BrBG(color_index / color_max)
for point in partition_dic[represent]:
fig.scatter(xs[point],
ys[point],
c=np.array([[r, g, b, a]]),
zorder=2)
color_index += 1
# Set limits :
fig.set_xlim(np.min(xs) - 0.1, np.max(xs) + 0.1)
fig.set_ylim(np.min(ys) - 0.1, np.max(ys) + 0.1)
def draw_perc_agw(ax, geom, vals, debug=False):
for g, val in zip(geom, vals):
if np.isfinite(val):
if debug:
x = g.centroid.x
y = g.centroid.y
ax.text(x, y, str(int(val)), transform=ccrs.PlateCarree())
quant = quantize_perc_agw(val)
facecolor = cm.BrBG(quant / 10)
else:
facecolor = 'grey'
ax.add_geometries([g],
ccrs.PlateCarree(),
facecolor=facecolor,
edgecolor=facecolor)
set_plot_parameters()
fig1 = plt.figure("pressure")
plt.clf()
#plt.subplots_adjust(right=0.75)
#plt.figtext(0.77,0.70, param_text,ha='left',va='top',size='x-small')
fig2 = plt.figure("density")
plt.clf()
# Loop thru particles
count = 0
for particle_id in particle_ids:
count = count + 1
particle_data = partextract(extract_exe, work_dir, uda_name, plot_dir,
particle_id)
plt_color = cm.BrBG(float(count) / float(len(particle_ids)))
plot_pressure(particle_data, particle_id, plt_color)
plot_density(particle_data, particle_id, plt_color)
plt.figure("pressure")
plt.grid(True)
plt.legend(loc='best', prop={'size': 10})
savePDF(fig1, 'pressure_vs_time', size='1280x960')
plt.figure("density")
plt.grid(True)
#plt.legend(loc='best', prop={'size':10})
savePDF(fig2, 'density_vs_time', size='1280x960')
plt.show()
import pandas as pd
import geopandas as gpd
import folium
import matplotlib.cm as cm
cuomo = 'ANDREW M. CUOMO'
nixon = 'CYNTHIA E. NIXON'
frac = 'cuomo_over_cuomo_plus_nixon'
def rgb_to_hex(rgb):
f = lambda x: int(x * 255)
return '#%02X%02X%02X' % (f(rgb[0]), f(rgb[1]), f(rgb[2]))
colormap = lambda x: rgb_to_hex(cm.BrBG(x))
df = pd.read_csv('20180913__ny__primary__nyc__precinct.csv')
gdf = gpd.read_file('shapefile/nyed.shp')
df['ElectDist'] = df['assembly district'].astype(
str) + df['election district'].apply(lambda x: str(x).zfill(3))
gov = df.pivot_table(index='ElectDist',
columns='candidate',
values='votes',
aggfunc=sum)[[cuomo, nixon]].astype(int)
gov[frac] = gov[cuomo] / (gov[cuomo] + gov[nixon])
gdf['ElectDist'] = gdf['ElectDist'].astype(str)
prob_ChR2=prob_ChR2,
frequency_A_ChR2=np.ascontiguousarray(frequency_A_ChR2),
ref_spectra_ChR2=np.ascontiguousarray(absorption_ChR2),
Raman_levels_ChR2=Raman_levels_ChR2,
levels_ChR2=levels_ChR2,
# GEVI molecule
energies_GEVI=energies_GEVI,
gamma_dep_GEVI=gamma_dep_GEVI,
prob_GEVI=prob_GEVI,
frequency_A_GEVI=np.ascontiguousarray(frequency_A_GEVI),
ref_spectra_GEVI=np.ascontiguousarray(absorption_GEVI),
Raman_levels_GEVI=Raman_levels_GEVI,
levels_GEVI=levels_GEVI)
colors = [cm.BrBG(x) for x in np.linspace(0., 1., M, endpoint=True)]
# ------------------------ SETTING UP PLOT GRIDS --------------------- #
# gs_top = plt.GridSpec(5, 2, top=0.95, hspace=0.2, wspace=0.05)
# gs_base = plt.GridSpec(5, 2, hspace=0.025, wspace=0.01)
# fig = plt.figure(figsize=(14, 11))
# axbig1 = fig.add_subplot(gs_top[:2, 0])
# axbig2 = fig.add_subplot(gs_top[:2, 1])
#
# axes = [[fig.add_subplot(gs_base[i, j]) for j in range(2)] for i in range(2, 5)]
#
# axbig2.set_yticklabels([])
# for i in range(2):
# axes[i][1].set_yticklabels([])
# for j in range(2):
drawbounds=True)
# collect the state names from the shapefile attributes so we can look up the shape obect for a state by it's name
names = []
colors = {}
i = 0
for shape_dict in m.states_info:
names.append(shape_dict['HASC_1'])
if shape_dict['HASC_1'] in corSTs.keys():
colors[
shape_dict['HASC_1']] = corSTs[shape_dict['HASC_1']] / 2 + 0.5
else:
colors[shape_dict['HASC_1']] = 0.5
ax = plt.gca() # get current axes instance
for nshape, seg in enumerate(m.states):
poly = Polygon(seg,
facecolor=cm.BrBG(colors[names[nshape]]),
edgecolor='k')
ax.add_patch(poly)
ARshp = m.readshapefile(
'/Volumes/Data_Archive/Data/adminBoundaries/GADM/ARG_adm_shp/ARG_adm1',
name='states',
drawbounds=True)
# collect the state names from the shapefile attributes so we can look up the shape obect for a state by it's name
names = []
colors = {}
i = 0
for shape_dict in m.states_info:
names.append(shape_dict['HASC_1'])
if shape_dict['HASC_1'] in corSTs.keys():
colors[
shape_dict['HASC_1']] = corSTs[shape_dict['HASC_1']] / 2 + 0.5