def individual_experiments(n, p, event):
"""
Print graphs with aproximated probabilities and its coutes
Input: list N, list P, int k
Output: -
"""
I = intervalo(0, n + 1, 1)
density = [0] * (len(I))
bundle = [0] * (len(I))
for i in I:
if event == 1:
density[i] = probability_last(n, p, i)
bundle[i] = bundle_last(n, p, i)
scatter_color = '#FF851B'
line_color = '#001f3f'
elif event == 2:
density[i] = probability_anyone(n, p, i)
scatter_color = '#FF851B'
bundle[i] = bundle_anyone(n, p, i)
line_color = '#FF4136'
else:
density[i] = probability_rigid_expansion(n, p, i)
scatter_color = '#eb7ab1'
bundle[i] = bundle_expansion(n, p, i)
line_color = '#85144b'
plt.plot(I,
density,
'o',
markersize=5,
alpha=0.8,
color=scatter_color,
label="Experimentation")
plt.plot(I,
bundle,
linewidth=1.5,
linestyle='-',
color=line_color,
label="Expected value")
plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left', borderaxespad=0.)
if event == 1:
plt.savefig('Figures/Single-uniquely-determinated-fixed-vertex.png',
bbox_inches="tight")
elif event == 2:
plt.savefig('Figures/Single-uniquely-determinated-any-vertex.png',
bbox_inches="tight")
else:
plt.savefig('Figures/Single-expansion-probability.png',
bbox_inches="tight")
plt.show()
def time_execution_performance(n, p, samples_size, log_scale=False):
"""
Plots a mean time rigid expansions experiments varing n, p and k.
Input: n (max graph size), p (list), samples_size(int)
Output: Mean time, float
"""
n_interval = intervalo(3,n+2,2)
k_proportion = [0.25, 0.5, 0.75]
# For the plotting
rows = len(k_proportion)
cols = len(p)
fig, axarr = plt.subplots(rows, cols, figsize=(3*cols+3,3*rows),dpi=80)
colors = color_template_3()
styles = plot_lines_styles()
cases = [
{"name":"Whitout optimizations", "optimizations":'none'},
{"name":"Fully optimized", "optimizations":"all"}]
for r in range(len(k_proportion)):
_proportion = k_proportion[r]
for c in range(len(p)):
_p = p[c]
for case in cases:
case_index = cases.index(case)
t = np.zeros(len(n_interval))
for i in range(len(n_interval)):
_n = n_interval[i]
_k = floor(_n * _proportion)
t[i] = calculate_mean_expansion_time(_n,
_p, _k, case, sample_size)
if log_scale:
t = log(t)
axarr[r,c].plot(n_interval, t, label=case['name'],
linestyle=styles[case_index%len(cases)], color=colors[case_index])
# Set labels for plot in row r and col c
print('Just finished', r,c)
axarr[r, c].spines['top'].set_color('white')
axarr[r, c].spines['right'].set_color('white')
axarr[r, c].set_title('p = ' + str(_p) + ', k_prop = ' + str(_proportion), backgroundcolor='silver')
axarr[r, c].set_ylabel('Time (ms)' + (' (log scale)' if log_scale else ''))
axarr[r, c].set_xlabel('Number of vertices')
if r == 0 and c == len(p)-1:
axarr[r,c].legend(bbox_to_anchor=(1.05, 1), loc='upper left', borderaxespad=0.)
fig.subplots_adjust(bottom=0.2)
plt.tight_layout()
plt.savefig('Figures/Time-execution-rigid-expansions'+('-log-scale' if log_scale else '')+'.png', bbox_inches='tight')
plt.show()
def graph_curves():
"""
Print bundles
Input: -
Output: -
"""
I = intervalo(5, 100, 1)
curve = [0] * (len(I))
properties = [{
'bundle': bundle_locally_infinite,
'color': '#ebc844',
'label': 'Locally infinite'
}, {
'bundle': bundle_connectivity,
'color': '#f16c20',
'label': 'Connected'
}, {
'bundle': bundle_diameter,
'color': '#0d3c55',
'label': 'Diameter 2'
}]
for prop in properties:
for i in range(0, I.size):
curve[i] = prop['bundle'](I[i])
plt.plot(I,
curve,
linewidth=1.5,
linestyle='-',
color=prop['color'],
label=prop['label'])
plt.legend(bbox_to_anchor=(1.05, 1),
loc='upper left',
borderaxespad=0.)
plt.fill_between(I, curve, 0.5, facecolor=prop['color'], alpha=0.4)
plt.savefig('Figures/ER-thresholds.png')
plt.show()
fig.colorbar(surf, shrink=0.5, aspect=5)
ax.w_xaxis.set_pane_color((1.0, 1.0, 1.0, 1.0))
ax.w_yaxis.set_pane_color((1.0, 1.0, 1.0, 1.0))
ax.w_zaxis.set_pane_color((1.0, 1.0, 1.0, 1.0))
plt.savefig(
'Figures/3D-Transition-matrix-secuence-of-rigid-expansions.png')
plt.show()
k = 50
sample = sampleJump(n, p, k, 1000)
m = mean(sample)
v = var(sample)
print(m / v)
n_est = m**2 / (m - v)
p_est = (m - v) / m
print(n_est)
print(p_est)
I = intervalo(min(sample), max(sample) + 1, 1)
density = [binomial(n_est, p_est, k) for k in I]
plt.hist(sample,
alpha=0.5,
facecolor='#cc0000',
edgecolor='#800000',
linewidth=1,
normed=1)
plt.plot(I, density, color="#800000", linewidth=1.5, label="densidad")
plt.title("Histogram")
plt.show()
def set_of_experiments(N, P, event):
"""
Print curves with aproximated probabilities and its cuotes
Input: list N, list P, int k
Output: -
"""
r, c = len(N), len(P)
goodness_of_fit = [[0] * c] * r
f, axarr = plt.subplots(r, c, figsize=(9, 6), dpi=100)
k = l = 0
for n in N:
I = intervalo(0, n + 1, 1)
density = [0] * (len(I))
bundle = [0] * (len(I))
for p in P:
for i in I:
if event == 1:
density[i] = probability_last(n, p, i)
bundle[i] = bundle_last(n, p, i)
scatter_color = '#0074d9'
line_color = '#001f3f'
elif event == 2:
density[i] = probability_anyone(n, p, i)
bundle[i] = bundle_anyone(n, p, i)
line_color = '#FF4136'
scatter_color = '#FF851B'
else:
density[i] = probability_rigid_expansion(n, p, i)
bundle[i] = bundle_expansion(n, p, i)
line_color = '#85144b'
scatter_color = '#eb7ab1'
goodness_of_fit[N.index(n)][P.index(p)] = '%.2E' % Decimal(
max([abs(x - y) for x, y in zip(density, bundle)]))
axarr[k, l].plot(I,
density,
'o',
markersize=5,
alpha=0.8,
color=scatter_color,
label="Experimentation")
axarr[k, l].plot(I,
bundle,
linewidth=1.5,
linestyle='-',
color=line_color,
label="Expected value")
#Boxes
if (k == 0 and l == 2):
axarr[k, l].legend(bbox_to_anchor=(1.05, 1),
loc='upper left',
borderaxespad=0.)
axarr[k, l].set_title('n=' + str(n) + ', p=' + str(p),
backgroundcolor='silver')
l = l + 1
l = 0
k = k + 1
#Pretty
f.subplots_adjust(hspace=0.4)
for i in range(0, r):
for j in range(0, c):
axarr[i, j].spines['top'].set_color('white')
axarr[i, j].spines['right'].set_color('white')
print("GOF", goodness_of_fit)
if event == 1:
plt.savefig('Figures/Uniquely-determinated-fixed-vertex.png',
bbox_inches="tight")
write_table("Uniquely-determinated-fixed-vertex-table-errors",
caption_goodness_of_fit, "gofExp1", N, P, goodness_of_fit)
plt.show()
elif event == 2:
plt.savefig('Figures/Uniquely-determinated-any-vertex.png',
bbox_inches="tight")
write_table("Uniquely-det-any-table-errors", caption_goodness_of_fit,
"gofExp2", N, P, goodness_of_fit)
plt.show()
else:
plt.savefig('Figures/Expansion-probability.png', bbox_inches="tight")
write_table("Expansion-probability-table-errors",
caption_goodness_of_fit, "gofExp3", N, P, goodness_of_fit)
plt.show()
plt.tight_layout()
return goodness_of_fit