def plot_time_selection(self):
arms = [ (k, v.p) for k,v in self.arms.iteritems() ]
sorted_arms = sorted(arms, key = lambda x: x[1])
data_sets = []
data_labels = []
for algo in self.pts.keys():
data = []
for pt in self.pts[algo]:
budget = sum([len(v) for v in pt.values()])
best = self._get_best(pt)
best_idx = [x[0] for x in sorted_arms].index(best)
data.append((budget, best_idx))
xvals = [pt[0] for pt in data]
yvals = [pt[1] for pt in data]
data_sets.append((xvals, yvals))
data_labels.append(algo)
make_plots.plot(data_sets, colors[:len(data_sets)],
[True for x in data_sets],
group_labels=data_labels,
plot_xlabel = 'Budget',
plot_ylabel = 'Selected Arm')
def plotLoss(train, val):
plot(
[[train[:, 1], train[:, 2]], [val[:, 1], val[:, 2]]],
series_colors=['green', 'red'],
series_labels=['Training', 'Validation'],
plot_ylim=[0, 9],
show_plot=False,
plot_xlabel='Number of Steps',
plot_ylabel='Loss',
#plot_title='Loss using Inception Net',
savefile='figs/{}_loss.pdf'.format(networkName))
def plotAccuracy(train, val):
plot(
[[train[:, 1], train[:, 2]], [val[:, 1], val[:, 2]]],
series_colors=['green', 'red'],
series_labels=['Training', 'Validation'],
plot_ylim=[0, 0.8],
show_plot=False,
plot_xlabel='Number of Steps',
plot_ylabel='Accuracy',
#savefile_size=(2.5, 1.5),
#plot_title='Accuracy using Inception Net',
savefile='figs/{}_accuracy.pdf'.format(networkName))
def plot_arm_dist(self, dist):
num_samples = 100000
if dist == 'uniform':
pvals = numpy.random.random_sample(num_samples)
elif dist == 'gauss-left':
pvals = numpy.random.normal(0.3, 0.1, num_samples)
elif dist == 'gauss-right':
pvals = numpy.random.normal(0.7, 0.1, num_samples)
edges = numpy.arange(0., 1., 0.01)
vals, edges = numpy.histogram(pvals, edges, density=True)
centers = [edges[i] + 0.5*(edges[i+1] - edges[i]) for i in range(len(edges)-1)]
make_plots.plot([(centers, vals)], ['grey'],
group_color_emphasis=[True],
plot_ylim=[0, max(vals)+0.01],
show_plot=True)
import scipy
import numpy as np
import ss_plotting.make_plots as ss
import matplotlib.pyplot as plt
data = np.loadtxt("./reproj_error.txt", delimiter=' ')
plt.plot(data)
plt.show()
rows = data.shape
t = range(0, rows[0])
ss.plot(series=[(t, data)],
series_colors=['blue'],
y_grid=True,
series_labels=[''],
plot_xlabel="Iterations",
plot_ylabel=" Median Reprojection Error (Pixels)",
savefile="./reproj_error.pdf",
savefile_size=(2, 2))
#plt.yscale('log')
plt.show()
import numpy as np
import ss_plotting.make_plots as ss
import matplotlib.pyplot as plt
data = np.loadtxt("./extrinsic_errors.txt", delimiter=' ')
plt.plot(data[:, 0:6] - data[:, 7:13])
plt.show()
rows = data.shape
# x y z x y z w x' y' z' w' x' y' z'
# 0 1 2 3 4 5 6 7 8 9 10 11 12 13
t = range(0, rows[0])
ss.plot(series=[(t, -(data[:, 0] - data[:, 7])),
(t, -(data[:, 1] - data[:, 8])),
(t, -(data[:, 2] - data[:, 9]))],
series_colors=['red', 'green', 'blue'],
series_labels=['X Error', 'Y Error', 'Z Error'],
plot_xlabel="Iteration",
plot_ylabel="Meters",
savefile="./extrinsic_error.pdf",
savefile_size=(3, 3),
y_grid=True)
#plt.yscale('log')
plt.show()
ss.plot(series=[(t, -(data[:, 3] - data[:, 10])),
(t, -(data[:, 4] - data[:, 11])),
(t, -(data[:, 5] - data[:, 12])),
(t, -(data[:, 6] - data[:, 13]))],
series_colors=['red', 'green', 'blue', 'black'],
series_labels=['QX Error', 'QY Error', 'QZ Error', 'QW Error'],
plot_xlabel="Iteration",
plot_ylabel="Radians",
import numpy as np
from ss_plotting.make_plots import plot
import matplotlib.pyplot as plt
data_translation = [
0.221, 0.306, 0.323, 0.337, 0.474, 0.205, 0.388, 0.499, 0.261, 0.625,
0.393, 0.508, 0.456
]
test_translation = [
0, 0, 0, 0, 0, 0.001, 0, 0.032, 0.024, 0.084, 0.081, 0.077, 0.075
]
x_trans = np.arange(0.65, 1.85, 0.1)
series = [(x_trans, data_translation), (x_trans, test_translation)]
series_labels = ['RGB', 'RGBD']
series_colors = ['red', 'blue']
ylabel = 'Error Percentage'
xlabel = 'Distance (cm)'
title = 'Rotation Error vs Distance with sigma = 0.5 pixels'
fig, ax = plot(series,
series_labels=series_labels,
series_colors=series_colors,
linewidth=5,
plot_ylabel=ylabel,
plot_xlabel=xlabel,
plot_title=title,
fontsize=13,
legend_fontsize=13)
0.252, 0.246, 0.254, 0.282, 0.262, 0.236, 0.204, 0.27, 0.262, 0.274, 0.256,
0.26, 0.284, 0.316, 0.27, 0.266, 0.294, 0.278, 0.27, 0.278, 0.298, 0.296,
0.302, 0.324, 0.322, 0.332, 0.3, 0.34, 0.336, 0.352, 0.318, 0.348, 0.368,
0.322, 0.328, 0.312, 0.326, 0.364, 0.378, 0.346, 0.342, 0.348, 0.332, 0.34,
0.37, 0.328, 0.36, 0.344, 0.38, 0.344, 0.324, 0.37, 0.372, 0.394, 0.384,
0.38, 0.36, 0.406, 0.4, 0.406, 0.346, 0.372, 0.372, 0.402, 0.358, 0.368,
0.418, 0.4, 0.368, 0.338, 0.39, 0.396, 0.404, 0.38, 0.406, 0.406, 0.414
]
x_trans = np.arange(0.6, 1.8, 0.01)
series = [(x_trans, data_translation)]
series_labels = ['Simulated Data']
series_colors = ['red']
fig, ax = plot(series,
series_labels=series_labels,
series_colors=series_colors,
linewidth=5)
data_rotation = [
0, 0, 0, 0, 0, 0, 0, 0, 0.008, 0.014, 0.046, 0.058, 0.112, 0.14, 0.182,
0.226, 0.328, 0.284, 0.276, 0.346, 0.322, 0.312, 0.36, 0.34, 0.31, 0.3,
0.298, 0.324, 0.296, 0.29, 0.274, 0.254, 0.286, 0.262, 0.254, 0.24, 0.224,
0.23, 0.208, 0.222, 0.216, 0.2, 0.182, 0.194, 0.168, 0.212, 0.178, 0.182,
0.196, 0.152, 0.174, 0.16, 0.16, 0.2, 0.168, 0.164, 0.154, 0.134, 0.156,
0.16, 0.13, 0.154, 0.134, 0.162, 0.144, 0.154, 0.114, 0.124, 0.126, 0.138,
0.144, 0.114, 0.104, 0.11, 0.098, 0.136, 0.098, 0.108, 0.096, 0.096, 0.116,
0.11, 0.12, 0.134, 0.132, 0.114, 0.126, 0.152, 0.136, 0.172
]
x_rot = np.arange(90)
series2 = [(x_rot, data_rotation)]
series_labels2 = ['Simulated Data']
]
data_trans_05 = [
0.0143, 0.0353, 0.0545, 0.0845, 0.1165, 0.1312, 0.1629, 0.1843, 0.216,
0.2415, 0.2641, 0.2692, 0.287, 0.3092, 0.3197, 0.3441, 0.3511, 0.3561,
0.3571, 0.3681, 0.3707, 0.3938, 0.397, 0.4062, 0.4131
]
data_trans_10 = [
0.1432, 0.1731, 0.2059, 0.2486, 0.2687, 0.3041, 0.3129, 0.3364, 0.3514,
0.3628, 0.3662, 0.3786, 0.3956, 0.4024, 0.4087, 0.4123, 0.4254, 0.4373,
0.4313, 0.4301, 0.4423, 0.4479, 0.4496, 0.4449, 0.4646
]
x_rot = np.arange(0.65, 1.85, 0.05)
series2 = [(x_rot, data_trans_02), (x_rot, data_trans_05),
(x_rot, data_trans_10)]
series_labels2 = ['sigma: 0.2 pixel', 'sigma: 0.5 pixel', 'sigma: 1.0 pixel']
series_colors2 = ['green', 'purple', 'orange']
ylabel = 'Error Percentage'
xlabel = 'Distance (m)'
title = 'Distance Error (Simulation)'
fig2, ax2 = plot(series2,
series_labels=series_labels2,
series_colors=series_colors2,
plot_ylabel=ylabel,
plot_xlabel=xlabel,
plot_title=title,
linewidth=5,
fontsize=15,
legend_fontsize=15)
plt.show()
import scipy
import numpy as np
import ss_plotting.make_plots as ss
import matplotlib.pyplot as plt
data = np.loadtxt("./position_error.txt", delimiter=' ')
plt.plot(data)
plt.show()
rows = data.shape
t = range(0, rows[0])
ss.plot(series=[(t, data[:, 0] - data[:, 3]), (t, data[:, 1] - data[:, 4]),
(t, data[:, 2] - data[:, 5])],
series_colors=['red', 'green', 'blue'],
y_grid=True,
series_labels=['X Error', 'Y Error', 'Z Error'],
plot_xlabel="Frame",
plot_ylabel="Meters",
savefile="./position_error.pdf",
savefile_size=(3, 3))
#plt.yscale('log')
plt.show()
#!bin/python
from IPython import embed
import numpy
import matplotlib.pyplot as plt
import scipy.stats as stats
import ss_plotting.make_plots as ss;
data = numpy.loadtxt(open("./dot_data.txt", "rb"), delimiter=' ');
theta = numpy.linspace(0, 3.14 * 2)
xes = numpy.sin(theta) * 2
yes = numpy.cos(theta) * 2
ss.plot(series=[(xes, yes), (data[:, 0], data[:, 1]), (data[:, 2], data[:, 3])],
series_colors=['red', 'blue', 'green'],
series_labels=['Dot Perimeter', 'Touches -- Dense Tracking', 'Touches -- Open Loop'],
plot_xlabel='X error (cm)',
plot_ylabel="Y error (cm)",
savefile="./dots.pdf",
line_styles=['-', '', ''],
plot_markers=['', '.', 'x'],
plot_xlim=(-8, 8),
plot_ylim=(-8, 8),
savefile_size = (3, 3));
plt.show();
def myfunc(row):
return numpy.sum(row) > 0.0001
print "Loading..."
data = numpy.loadtxt('./offsets.txt', delimiter=' ')
#data = data[numpy.array([myfunc(row) for row in data])]
m = numpy.max(numpy.abs(data))
t = xrange(1, 200)
ss.plot(series=[(t, data[1:200, 1]), (t, data[1:200, 2])],
series_colors=['blue', 'red'],
y_grid=True,
series_labels=['$q_2$', '$q_3$'],
plot_xlabel="Frame",
plot_ylabel="Offset (rad)",
savefile="./kinematic_redundancy.pdf",
savefile_size=(4, 3))
plt.plot(t, data[1:200, 1], '-')
plt.plot(t, data[1:200, 2], '-')
plt.show()
f, axarr = plt.subplots(3, 2, sharex=True)
for i in xrange(0, 6):
ax = axarr[i / 2, i % 2]
ax.plot(data[:, i], '.', color=(0.6, 0.0, 0.0))
ax.plot(data[:, i + 6], '.', color=(0.6, 0.6, 0.6))
#ax.plot(smooth(data[:, i], 9), '-', color=(0, 0, 0));
import numpy as np;
import ss_plotting.make_plots as ss;
import matplotlib.pyplot as plt;
import os
datas = dict();
ts = [];
errs = [];
for dirname, dirnames, filenames in os.walk('.'):
dirnames = sorted(dirnames)
# print path to all subdirectories first.
for subdirname in dirnames:
print subdirname
datas[subdirname] = np.loadtxt(os.path.join(dirname, subdirname) + "/extrinsic_error.txt")
ts.append(float(subdirname))
errs.append(50 * np.sqrt(np.dot(np.transpose(datas[subdirname][7:10]), datas[subdirname][7:10])));
# print path to all filenames.
#for filename in filenames:
# print(os.path.join(dirname, filename))
ss.plot(series=[(ts, errs)], series_colors=['blue'],
series_labels=[''], plot_xlabel="Camera field of view (degrees)", plot_ylabel="Final Extrinsic Error (cm)", savefile="./extrinsic_fov.pdf", savefile_size=(3, 3), y_grid=True,
line_styles=[''], plot_markers=['.']);
##print errs
##plt.plot(ts, errs, 'x')
##plt.show();
group_color_emphasis = [True],
group_labels = None,
plot_ylabel = 'Counts',
plot_xlabel = 'Success Probability',
bin_labels = bin_edges + 0.5*bin_size,
savefile = savefile,
savefile_size = (.5*4.75, .5*.5*4.75))
# Add expected value data
all_data.append((xvals, expected_vals))
all_labels.append('Expected - %s' % arm_dist)
all_colors.append(colors[aidx])
# all_color_emphasis.append(False)
all_color_emphasis.append(True)
# Add actual value data
# all_data.append((xvals, actual_vals))
# all_labels.append('Actual - %s' % arm_dist)
# all_colors.append(colors[aidx])
# all_color_emphasis.append(True)
savefile = 'comparison.png' if args.save_plots else None
make_plots.plot(all_data, all_colors,
group_color_emphasis = all_color_emphasis,
group_labels = all_labels,
plot_xlabel = 'Num selected arms',
plot_ylabel = 'Success probability',
fontsize=8, legend_fontsize=8,
savefile = savefile,
savefile_size = (.5*4.75, 1.5*.5*4.75))
