results["title"].append(title)
print
def cm2inch(value):
return value/2.54
def plot():
fig = plt.figure(figsize=(cm2inch(13), cm2inch(6)))
ax = fig.add_subplot(111)
bp = ax.boxplot(results["data"])
plt.setp(bp['boxes'], color='black')
plt.setp(bp['whiskers'], color='black')
plt.setp(bp['fliers'], color='black', marker='+')
plt.setp(bp['medians'], color='black')
ax.set_xticklabels(results["title"])
ax.set_ylim(0, np.max(results["data"]))
ax.set_ylabel("Occupancy")
plt.savefig('out/binam_occupancy_data.pdf', format='pdf',
bbox_inches='tight')
calculate(lambda: binam_data.generate_random(n_bits, n_ones, n_samples),
"Random\n(with duplicates)")
calculate(lambda: binam_data.generate(n_bits, n_ones, n_samples, balance=False),
"Random\n(no duplicates)")
calculate(lambda: binam_data.generate_naive(n_bits, n_ones, n_samples),
"Balanced\n(with duplicates)")
calculate(lambda: binam_data.generate(n_bits, n_ones, n_samples),
"Balanced\n(no duplicates)")
plot()
print "Calculating correlation for \"", title, "\""
C = np.corrcoef(m.T)
print "Plotting..."
fig = plt.figure(figsize=(cm2inch(7.5), cm2inch(6)))
ax = fig.add_subplot("111")
ax.set_xlabel("Bit index $i$")
ax.set_ylabel("Bit index $j$")
ax.set_title(title)
cax = ax.imshow(C, interpolation="none", vmin=0, vmax=0.1,
cmap="Blues")
fig.colorbar(cax, ticks=[0, 0.05, 0.1])
return fig
# Plot the own data
print "Generate own data...."
plot(binam_data.generate(size, nBits, nSamples),
"With selection bias")\
.savefig("out/balanced_with_bias.pdf", format='pdf', bbox_inches='tight')
print "Generate own data (with weight_choices=False)..."
plot(binam_data.generate(size, nBits, nSamples, weight_choices=False),
"Without selection bias")\
.savefig("out/balanced_no_bias.pdf", format='pdf', bbox_inches='tight')
#print "Generate own data (naive method)..."
#plot(binam_data.generate_naive(size, 3, 10000),
# "Python generator (naive)")
plt.show()
def cm2inch(value):
return value / 2.54
def plot():
fig = plt.figure(figsize=(cm2inch(13), cm2inch(6)))
ax = fig.add_subplot(111)
bp = ax.boxplot(results["data"])
plt.setp(bp['boxes'], color='black')
plt.setp(bp['whiskers'], color='black')
plt.setp(bp['fliers'], color='black', marker='+')
plt.setp(bp['medians'], color='black')
ax.set_xticklabels(results["title"])
ax.set_ylim(0, np.max(results["data"]))
ax.set_ylabel("Occupancy")
plt.savefig('out/binam_occupancy_data.pdf',
format='pdf',
bbox_inches='tight')
calculate(lambda: binam_data.generate_random(n_bits, n_ones, n_samples),
"Random\n(with duplicates)")
calculate(
lambda: binam_data.generate(n_bits, n_ones, n_samples, balance=False),
"Random\n(no duplicates)")
calculate(lambda: binam_data.generate_naive(n_bits, n_ones, n_samples),
"Balanced\n(with duplicates)")
calculate(lambda: binam_data.generate(n_bits, n_ones, n_samples),
"Balanced\n(no duplicates)")
plot()
print "Calculating correlation for \"", title, "\""
C = np.corrcoef(m.T)
print "Plotting..."
fig = plt.figure(figsize=(cm2inch(7.5), cm2inch(6)))
ax = fig.add_subplot("111")
ax.set_xlabel("Bit index $i$")
ax.set_ylabel("Bit index $j$")
ax.set_title(title)
cax = ax.imshow(C, interpolation="none", vmin=0, vmax=0.1, cmap="Blues")
fig.colorbar(cax, ticks=[0, 0.05, 0.1])
return fig
# Plot the own data
print "Generate own data...."
plot(binam_data.generate(size, nBits, nSamples),
"With selection bias")\
.savefig("out/balanced_with_bias.pdf", format='pdf', bbox_inches='tight')
print "Generate own data (with weight_choices=False)..."
plot(binam_data.generate(size, nBits, nSamples, weight_choices=False),
"Without selection bias")\
.savefig("out/balanced_no_bias.pdf", format='pdf', bbox_inches='tight')
#print "Generate own data (naive method)..."
#plot(binam_data.generate_naive(size, 3, 10000),
# "Python generator (naive)")
plt.show()
