import os, sys
import cPickle
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import matplotlib.image as image
# import tqdm
### use mpl 1.0 style
import matplotlib as mpl
mpl.style.use('classic')
import seaborn.apionly as sns
colors = sns.color_palette("colorblind")
sns.set_palette(sns.color_palette("colorblind"))
from keplerian import keplerian
from display import DisplayResults
# sys.path.append('/home/joao/Work/OPEN')
# from OPEN.ext.keplerian import keplerian
sys.path.append('.')
from styler import styler
# res = DisplayResults('')
# res.sample = np.atleast_2d(np.loadtxt('sample.txt'))
res = cPickle.load(open('BL2009_joss_figure1.pickle'))
from __future__ import division
import argparse
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn.apionly as sns
from sklearn.model_selection import learning_curve
from composition.analysis.load_sim import load_sim
from composition.analysis.preprocessing import get_train_test_sets
from composition.analysis.pipelines import get_pipeline
if __name__ == '__main__':
sns.set_palette('muted')
sns.set_color_codes()
p = argparse.ArgumentParser(
description='Runs extra modules over a given fileList')
p.add_argument('-c',
'--classifier',
dest='classifier',
default='RF',
choices=['RF', 'KN'],
help='Option to specify classifier used')
p.add_argument('--outdir',
dest='outdir',
default='/home/jbourbeau/public_html/figures/composition',
help='Output directory')
args = p.parse_args()
import numpy as np
import os
import seaborn.apionly as sns
sns.set_palette('muted')
cls = (sns.color_palette())
import matplotlib.pyplot as plt
save_folder = '/Users/Adriaan/GitHubRepos/DiCarloLab_Repositories/Paper_16_RestlessTuning/Figures'
save_format = 'pdf'
def make_figure(fig_name='timing_tech',
Acquisition_time=np.array([1.6, 0.122, 0.122]),
AWG_overhead=np.array([0.093, 0.093, 0]),
Processing_overhead=np.array([.164+.063, .164+.063, 0]),
Overhead=np.array([0.04, 0.04, 0.04]),
methods=('Conventional-', 'Restless-', 'Restless+'),
):
cls = (sns.color_palette('muted'))
plt.rcParams.update({'font.size': 9, 'legend.labelspacing': 0,
'legend.columnspacing': .3,
'legend.handletextpad': .2})
f, ax = plt.subplots(figsize=(3.3, .9))
y_pos = np.array([.8, 0, -.8])+.1
lefts = np.array([0, 0, 0])
ax.barh(y_pos, AWG_overhead, color=cls[2], align='center',
height=0.6, label='Set pars.')
lefts = lefts+np.array(AWG_overhead)
import numpy as np
import pandas as pd
import torch
import pickle
from tqdm import tqdm_notebook as tqdm
import statsmodels.formula.api as smf
import statsmodels.api as sm
import seaborn.apionly as sns
import matplotlib
import matplotlib.pyplot as plt
plt.style.use('clean')
sns.set_palette(['#8F223A', '#505B70', '#A6A29F'])
plt.rcParams['font.sans-serif'] = 'Verdana'
plt.rcParams['figure.figsize'] = (6.5, 4)
%matplotlib inline
%config InlineBackend.figure_format = 'retina'
def ft_ax(ax=None,
y=1.03,
yy=1.1,
title=None,
subtitle=None,
source=None,
add_box=False,
left_axis=False):
"""
def plot_joint_moment_breakdown(time,
joint_moments,
tendon_forces,
moment_arms,
dof_names,
muscle_names,
pdf_path,
csv_path,
ext_moments=None,
ext_names=None,
ext_colors=None,
mass=None):
"""Plots net joint moments, individual muscle moments, and, if included,
any external moments in the system. Prints a pdf file with plots and a csv
file containing data.
N: # of time points
Ndof: # of degrees-of-freedom
Nmusc: # of muscles
Parameters
----------
time (N,): Numpy array
Vector of time points
joint_moments (N, Ndof): Numpy array
Array of experimental net joint moments.
tendon_forces (N, Nmusc): Numpy array
Array of computed tendon forces.
moment_arms (N, Ndof, Nmusc): Numpy array
Array of muscle moment arms, corresponding to each joint.
dof_names (Ndof,): list
List of strings containing degree-of-freedom names.
muscle_names (Nmusc,): list
List of strings containing muscle names.
pdf_path: string
Path and filename for PDF of final plots.
csv_path: string
Path and filename for CSV of moment breakdown data
mass: kg
(Optional). Subject mass to normalize moments by.
ext_moments:
(Optional). External moments applied to each degree-of-freedom (i.e.
such as an exoskeleton device torque).
"""
num_dofs = len(dof_names)
num_muscles = len(muscle_names)
# For writing moments to a file.
dof_array = list()
actuator_array = list()
moments_array = list()
all_moments = list()
import seaborn.apionly as sns
palette = sns.color_palette('muted')
num_colors = 6
sns.set_palette(palette)
pgc = 100.0 * (time - time[0]) / (time[-1] - time[0])
pgc_csv = np.linspace(0, 100, 400)
fig = pl.figure(figsize=(8.5, 11))
for idof in range(num_dofs):
dof_name = dof_names[idof]
ax = fig.add_subplot(num_dofs, 2, 2 * idof + 1)
net_integ = np.trapz(np.abs(joint_moments[:, idof]), x=time)
sum_actuators_shown = np.zeros_like(time)
icolor = 0
for imusc in range(num_muscles):
muscle_name = muscle_names[imusc]
if np.any(moment_arms[:, idof, imusc]) > 0.00001:
this_moment = \
tendon_forces[:, imusc] * moment_arms[:, idof, imusc]
mom_integ = np.trapz(np.abs(this_moment), time)
if mom_integ > 0.05 * net_integ:
if np.floor(icolor / num_colors) == 0:
ls = '-'
elif np.floor(icolor / num_colors) == 1:
ls = '--'
else:
ls = '-.'
ax.plot(pgc, this_moment, label=muscle_name, linestyle=ls)
dof_array.append(dof_name)
actuator_array.append(muscle_name)
all_moments.append(np.interp(pgc_csv, pgc, this_moment))
sum_actuators_shown += this_moment
icolor += 1
if ext_moments:
num_ext = len(ext_moments)
for iext in range(num_ext):
ext_moment = ext_moments[iext]
if ext_names:
ext_name = ext_names[iext]
if ext_colors:
ext_color = ext_colors[iext]
pgc_mod = 100.0 * (
(ext_moment.index - ext_moment.index[0]) * 1.0 /
(ext_moment.index[-1] - ext_moment.index[0]) * 1.0)
ext_col = -1
for colname in ext_moment.columns:
if colname in dof_name:
ext_col = colname
break
if ext_col == -1:
raise Exception('Could not find exo torque for DOF %s.' %
dof_name)
if np.sum(np.abs(ext_moment[ext_col])) != 0:
sum_actuators_shown += ext_moment[ext_col]
ax.plot(pgc_mod,
ext_moment[ext_col],
color=ext_color if ext_color else 'blue',
label=ext_name if ext_name else 'external',
ls='dashed',
linewidth=1.5)
dof_array.append(dof_name)
actuator_array.append(ext_name if ext_name else 'external')
all_moments.append(
np.interp(pgc_csv, pgc_mod, ext_moment[ext_col]))
ax.plot(pgc,
sum_actuators_shown,
label='sum actuators shown',
color='gray',
linewidth=2)
ax.plot(pgc,
joint_moments[:, idof],
label='net',
color='black',
linewidth=2)
dof_array.append(dof_name)
actuator_array.append('net')
all_moments.append(np.interp(pgc_csv, pgc, joint_moments[:, idof]))
ax.set_title(dof_name, fontsize=8)
ax.set_ylabel('moment (N-m)', fontsize=8)
ax.legend(frameon=False,
bbox_to_anchor=(1, 1),
loc='upper left',
ncol=2,
fontsize=8)
ax.tick_params(axis='both', labelsize=8)
ax.set_xlabel('time (% gait cycle)', fontsize=8)
fig.tight_layout()
fig.savefig(pdf_path)
pl.close(fig)
multiindex_arrays = [dof_array, actuator_array]
columns = pd.MultiIndex.from_arrays(multiindex_arrays,
names=['DOF', 'actuator'])
# Normalize by subject mass.
if mass:
all_moments_array = (np.array(all_moments).transpose() / mass)
moments_df = pd.DataFrame(all_moments_array,
columns=columns,
index=pgc_csv)
with file(csv_path, 'w') as f:
f.write('# all columns are moments normalized by subject '
'mass (N-m/kg).\n')
moments_df.to_csv(f)
