# vim:fileencoding=utf-8
import matplotlib.pyplot as plt
import numpy as np
import freqanalysis.ecdf as ecdf
import freqanalysis.datatools as datatool
from scipy.stats import ks_2samp
datasetfile = "datasets/20140904-export.txt"
print "loading ", datasetfile
df = datatool.load_data_as_dataframe(datasetfile)
print "Calculating ECDF of all values"
all_series, yvals = ecdf.get_ecdf(df['freq'])
print "Plotting graph"
ecdf.plot_ecdf_curve(all_series, yvals, color="b", label="Alle Werte")
df['minute'] = df.time.apply(lambda x: x.minute)
hour_df = df[(df.minute >= 58) | (df.minute <= 5)]
hour_series, yvals = ecdf.get_ecdf(hour_df['freq'])
ecdf.plot_ecdf_curve(hour_series, yvals, color="r",
label="Stundenwechsel")
not_hour_df = df[(df.minute < 58 ) & (df.minute > 5)]
not_hour_series, yvals = ecdf.get_ecdf(not_hour_df['freq'])
ecdf.plot_ecdf_curve(not_hour_series, yvals, color="y", linestyle="-",
label="unter der Stunde")
print "Null hypothesis: the two samples are drawn from the same continuous distribution."
D, p_value = ks_2samp(all_series, hour_series)
cp.set_label(u"Häufigkeit")
#plt.plot(eclipse_momentum_df.s_since_midnight.astype(int),
# eclipse_momentum_df.momentum, 'r-', label="Momentum")
#hfmt = dates.DateFormatter('%H:%M')
#ax.xaxis.set_major_formatter(hfmt)
# y_formatter = mpl.ticker.ScalarFormatter(useOffset=False)
# ax.yaxis.set_major_formatter(y_formatter)
# ax.grid(True)
f.suptitle("Dichte der Leistungsgradienten")
f.autofmt_xdate()
plt.savefig("images/sonnenfinsternis-dichte-gradienten.png")#, bbox_inches='tight')
plt.clf()
friday_series, friday_vals = ecdf.get_ecdf(friday_momentum_df.momentum)
ecdf.plot_ecdf_curve(friday_series, friday_vals, color="b", label="Typischer Freitag")
eclipse_series, eclipse_vals = ecdf.get_ecdf(eclipse_momentum_df.momentum)
ecdf.plot_ecdf_curve(eclipse_series, eclipse_vals, color="r", label="Sonnenfinsternis")
print "Mittelwert alle Freitage: %f" % np.median(friday_momentum_df.momentum)
print "Mittelwert Sonnenfinsternis: %f" % np.median(eclipse_momentum_df.momentum)
# http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.levene.html#scipy.stats.levene
W, p_val = stats.levene(friday_momentum_df.momentum,
eclipse_momentum_df.momentum, center='median')
print ("Levenes Test auf Gleichheit der Varianz: P=%s (gleiche Varianz für p<=0.05)" % p_val)
W, p_val = stats.fligner(friday_momentum_df.momentum, eclipse_momentum_df.momentum)
print "Fliegners Test auf Gleichheit der Varianz: P=%s" % p_val
f.suptitle("ECDF der Leistungsgradienten: Ungleiche Varianzen (Levene, p=%f)" % p_val)
#plt.plot(eclipse_momentum_df.s_since_midnight.astype(int),
# eclipse_momentum_df.momentum, 'r-', label="Momentum")
#hfmt = dates.DateFormatter('%H:%M')
#ax.xaxis.set_major_formatter(hfmt)
# y_formatter = mpl.ticker.ScalarFormatter(useOffset=False)
# ax.yaxis.set_major_formatter(y_formatter)
# ax.grid(True)
f.suptitle("Dichte der Leistungsgradienten")
f.autofmt_xdate()
plt.savefig("images/sonnenfinsternis-dichte-gradienten.png"
) #, bbox_inches='tight')
plt.clf()
friday_series, friday_vals = ecdf.get_ecdf(friday_momentum_df.momentum)
ecdf.plot_ecdf_curve(friday_series,
friday_vals,
color="b",
label="Typischer Freitag")
eclipse_series, eclipse_vals = ecdf.get_ecdf(eclipse_momentum_df.momentum)
ecdf.plot_ecdf_curve(eclipse_series,
eclipse_vals,
color="r",
label="Sonnenfinsternis")
print "Mittelwert alle Freitage: %f" % np.median(
friday_momentum_df.momentum)
print "Mittelwert Sonnenfinsternis: %f" % np.median(
eclipse_momentum_df.momentum)
# http://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.levene.html#scipy.stats.levene
W, p_val = stats.levene(friday_momentum_df.momentum,
