num_subjects = 8
bar_color = [0.75, 0.75, 0.75]
x_pos = np.arange(0, 2)
scatter_pts = np.linspace(-0.1, 0.1, num_subjects)
phase_err_data = np.concatenate(
(data['phase_ekf_err_mean'], data['phase_time_err_mean']), 1)
phase_err_median = np.median(phase_err_data, 0)
#add num falls
ax.bar(x_pos,
phase_err_median,
color=bar_color,
tick_label=('GP-EKF', 'Time-Based'))
add_barplot_sigstars(ax, np.array([(0, 1)]), data['p_values'].flatten(), x_pos)
scatter_opts = {'s': 24, 'zorder': 10}
subject_list = np.concatenate((np.arange(1, 8), [0]))
marker_able = 'o'
marker_exp = 's'
for i in range(2):
for sub in subject_list:
if sub == 0:
marker = marker_exp
else:
marker = marker_able
ax.scatter(x_pos[i] + scatter_pts[sub],
phase_err_data[sub, i],
marker=marker,
ax.set_ylabel('Number of Falls')
num_subjects = 10
bar_color = [0.75, 0.75, 0.75]
x_pos = np.array((0, 1, 2, 3, 5, 6, 7, 8))
scatter_pts = np.linspace(-0.1, 0.1, num_subjects / 2)
ax.bar(x_pos,
np.median(data['fall_counts_subopt'], 0),
color=bar_color,
tick_label=('NM', 'NM\nsubopt', 'IMP', 'IMP\nsubopt', 'NM',
'NM\nsubopt', 'IMP', 'IMP\nsubopt'))
add_barplot_sigstars(ax,
data['condition_combinations_subopt'] - 1,
data['p_values_falls_subopt'],
x_pos,
star_loc='level')
subject_list = range(num_subjects)
scatter_opts = {'s': 10, 'zorder': 10, 'marker': 'o'}
for i in range(8):
if any([i == elem for elem in [0, 1, 4, 5]]):
colors_sub = colors[1:num_subjects + 2:2]
else:
colors_sub = colors[0:num_subjects + 2:2]
ax.scatter(x_pos[i] + scatter_pts,
data['fall_counts_subopt'][:, i],
color=colors_sub,
**scatter_opts)
fig, ax = plt.subplots(1, 1, figsize=(4.5, 2))
ax.set_ylabel('Number of Falls')
num_subjects = 10
bar_color = [0.75, 0.75, 0.75]
x_pos = np.array((0, 1, 2, 4, 5, 6))
scatter_pts = np.linspace(-0.1, 0.1, num_subjects)
ax.bar(x_pos,
np.median(data['fall_counts'], 0),
color=bar_color,
tick_label=('No Pros', 'NM', 'IMP', 'No Pros', 'NM', 'IMP'))
add_barplot_sigstars(ax,
data['condition_combinations'] - 1,
data['p_values_falls'],
x_pos,
star_loc='3x3')
subject_list = range(num_subjects)
scatter_opts = {'s': 10, 'zorder': 10, 'marker': 'o'}
for i in range(6):
ax.scatter(x_pos[i] + scatter_pts,
data['fall_counts'][:, i],
color=colors,
**scatter_opts)
trans = transforms.blended_transform_factory(ax.transData, ax.transAxes)
group_label_props = {
'horizontalalignment': 'center',
'verticalalignment': 'center',
fig, ax = plt.subplots(1, 1, figsize=(2, 2))
ax.set_ylabel(r"Ankle Net Work (\unitfrac{J}{kg})")
num_subjects = 10
bar_color = [0.75, 0.75, 0.75]
x_pos = np.array((0, 1))
scatter_pts = np.linspace(-0.1, 0.1, num_subjects)
ax.bar(x_pos,
np.mean(data['ankle_net_work'][:, 1:], 0),
color=bar_color,
tick_label=('NM', 'IMP'))
add_barplot_sigstars(ax,
np.array([[0, 1]]),
np.array([data['p_values_ankle_net_work'][2]]),
x_pos,
star_loc='level')
subject_list = range(num_subjects)
scatter_opts = {'s': 10, 'zorder': 10, 'marker': 'o'}
for i in range(2):
ax.scatter(x_pos[i] + scatter_pts,
data['ankle_net_work'][:, i + 1],
color=colors,
**scatter_opts)
trans = transforms.blended_transform_factory(ax.transData, ax.transAxes)
group_label_props = {
'horizontalalignment': 'center',
'verticalalignment': 'center',
fig, ax = plt.subplots(1, 1, figsize=(4.5, 3))
ax.set_ylabel('Step Length\nVariability (mm)')
num_subjects = 10
bar_color = [0.75, 0.75, 0.75]
x_pos = np.array((0, 1, 2, 4, 5, 6))
scatter_pts = np.linspace(-0.1, 0.1, num_subjects)
ax.bar(x_pos,
np.mean(data['step_length_var'], 0),
color=bar_color,
tick_label=('No Pros', 'NM', 'IMP', 'No Pros', 'NM', 'IMP'))
add_barplot_sigstars(ax,
data['condition_combinations'] - 1,
data['p_values_step_length_var'],
x_pos,
star_loc='3x3')
subject_list = range(num_subjects)
scatter_opts = {'s': 10, 'zorder': 10, 'marker': 'o'}
for i in range(6):
ax.scatter(x_pos[i] + scatter_pts,
data['step_length_var'][:, i],
color=colors,
**scatter_opts)
trans = transforms.blended_transform_factory(ax.transData, ax.transAxes)
group_label_props = {
'horizontalalignment': 'center',
'verticalalignment': 'center',
num_subjects = 9
bar_color = [0.75, 0.75, 0.75]
x_pos = np.arange(0,3)
scatter_pts = np.linspace(-0.1,0.1,num_subjects)
scatter_opts = {'s':8, 'zorder':10}
subject_list = np.concatenate((np.arange(1,8), [0], [8]))
marker_able = 'o'
marker_amp = '^'
marker_exp = 's'
#plot bar plots and sig stars
ax[0,0].bar(x_pos, 180/np.pi*data['knee_angle_errors_median'].flatten(),
color = bar_color)
add_barplot_sigstars(ax[0,0], data['condition_combinations']-1,
data['p_values_knee_angle'].flatten(), x_pos, star_loc='3x1')
ax[0,1].bar(x_pos, 180/np.pi*data['ankle_angle_errors_median'].flatten(),
color = bar_color)
add_barplot_sigstars(ax[0,1], data['condition_combinations']-1,
data['p_values_ankle_angle'].flatten(), x_pos, star_loc='3x1')
ax[1,0].bar(x_pos, data['knee_moment_errors_median'].flatten(),
color = bar_color, tick_label=('GP-EKF','NM','IMP'))
add_barplot_sigstars(ax[1,0], data['condition_combinations']-1,
data['p_values_knee_moment'].flatten(), x_pos, star_loc='3x1')
ax[1,1].bar(x_pos, data['ankle_moment_errors_median'].flatten(),
color = bar_color, tick_label=('GP-EKF','NM','IMP'))
add_barplot_sigstars(ax[1,1], data['condition_combinations']-1,
data['p_values_ankle_moment'].flatten(), x_pos, star_loc='3x1')
ax.set_ylabel('Phase Transition\nSuccess Rate')
num_subjects = 10
bar_color = [0.75, 0.75, 0.75]
x_pos = np.array((0, 1))
scatter_pts = np.linspace(-0.1, 0.1, num_subjects)
ax.bar(x_pos,
np.mean(data['phase_success'], 0),
color=bar_color,
tick_label=('No Disturb', 'w/ Disturb'))
pdb.set_trace()
add_barplot_sigstars(ax,
np.array([data['condition_combinations'] - 1]),
np.array([data['p_values_phase_success']]),
x_pos,
star_loc='level')
subject_list = range(num_subjects)
scatter_opts = {'s': 10, 'zorder': 10, 'marker': 'o'}
for i in range(2):
ax.scatter(x_pos[i] + scatter_pts,
data['phase_success'][:, i],
color=colors,
**scatter_opts)
trans = transforms.blended_transform_factory(ax.transData, ax.transAxes)
group_label_props = {
'horizontalalignment': 'center',
'verticalalignment': 'center',
