def run_experiment(tweets_csv, keywords, timezoneslist, titles, colors):
for timezones, title in zip(timezoneslist, titles):
fig = plt.figure(figsize=(6,6))
ax = plt.axes()
loc = MultipleLocator(base=5.0)
loc.MAXTICKS = 10000
ax.xaxis.set_major_formatter(xfmt)
ax.xaxis.set_major_locator(loc)
box = ax.get_position()
#ax.set_position([box.x0, box.y0, box.width*0.85, box.height])
if len(timezones) == 1 and timezones[0] == "":
ax.set_title("All timezones")
else:
ax.set_title("Timezone: {}".format(title))
x, y, labels = foreach_keyword(tweets_csv, keywords, timezones)
xf, yf = fill_with_zeros(x, y)
plot_stacked_timelines(ax, xf, yf, colors, labels)
ax.set_xlabel("Date")
ax.set_ylabel("Tweet count")
ax.xaxis_date()
ax.autoscale()
#plt.legend(loc="best", bbox_to_anchor=(1, 0.6))
plt.savefig("frequency_stacked_{}.pdf".format(title))
#!/usr/bin/python
import sys
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import pandas
from matplotlib.ticker import MultipleLocator, FormatStrFormatter
plt.clf()
fig=plt.figure(figsize=(40, 40))
ax = fig.add_subplot(111)
mat=pandas.read_csv(sys.argv[1],header=0,index_col=0)
# min=0 and max=1 is set to just distinguish absent and present
# remove those parameters to see colors scaled by value
ax.matshow(mat,cmap=cm.gray, interpolation='nearest', vmin=0, vmax=1)
locator=MultipleLocator(1)
locator.MAXTICKS = 10000
ax.xaxis.set_major_locator(locator)
ax.set_xticklabels([""]+mat.columns.tolist(), rotation=90)
ax.yaxis.set_major_locator(MultipleLocator(1))
ax.yaxis.set_major_formatter(FormatStrFormatter('%s'))
ax.set_yticklabels([""]+mat.index.tolist())
ax.tick_params(axis='both', which='both', labelsize=8, direction='out',
labelleft='on', labelright='off', labelbottom='off',
labeltop='on', left='on', right='off', bottom='off',
top='on')
plt.savefig(sys.argv[2], bbox_inches='tight', dpi = 100)
