def _get_te(blocks):
task_entries = []
for block in blocks:
if np.iterable(block):
task_entries.append(CombinedTaskEntry(block))
else:
task_entries.append(performance._get_te(block))
return task_entries
def test_block(idx, n_iter=None):
te = performance._get_te(idx)
if n_iter == None:
n_iter = len(te.hdf.root.task)
gen = genfns.sim_target_seq_generator_multi(8, 1000)
task = tasks.BMIReconstruction(te, n_iter, gen)
task.init()
error = task.calc_recon_error(verbose=False)
abs_max_error = np.max(np.abs(error))
return abs_max_error
def test_block(idx, n_iter=None):
te = performance._get_te(idx)
if n_iter == None:
n_iter = len(te.hdf.root.task)
gen = genfns.sim_target_seq_generator_multi(8, 1000)
task = tasks.BMIReconstruction(te, n_iter, gen)
task.init()
error = task.calc_recon_error(verbose=False)
abs_max_error = np.max(np.abs(error))
return abs_max_error
def fn(task_entry_model):
te = performance._get_te(task_entry_model)
try:
return te.n_trials >= min_trial_count
except:
return False
def _check_decoder(task_entry_model):
te = performance._get_te(task_entry_model)
try:
return (decoder_type in str(te.decoder_type))
except:
return False
from tasks import performance import matplotlib.pyplot as plt from riglib.bmi import robot_arms, assist, train, goal_calculators from riglib.stereo_opengl import ik import time from riglib import mp_calc reload(mp_calc) reload(assist) reload(robot_arms) reload(goal_calculators) pi = np.pi te = performance._get_te(3086) target = te.hdf.root.task[:]['target'] starting_pos_ws = np.array([5., 0, 5]) shoulder_anchor = np.array([2, 0, -15]) # Generate the list of targets from tasks import generatorfunctions target_list = generatorfunctions.centerout_2D_discrete(nblocks=1, ntargets=2) target_list = target_list.reshape(-1, 3) n_targets = target_list.shape[0] # Initialize the kinematic chain chain = robot_arms.PlanarXZKinematicChain([15, 15, 5, 5]) chain.joint_limits = [(-pi, pi), (-pi, 0), (-pi / 2, pi / 2),
#!/usr/bin/python
'''
Test that the content of the npz file is the same as what is stored in the HDF file
'''
from db import dbfunctions as dbfn
from db.tracker import models
from tasks import performance
import plotutil
import tables
import numpy as np
te = performance._get_te(2440)
print te
te.clda_param_hist
error = np.zeros(len(te.clda_param_hist))
for k, update in enumerate(te.clda_param_hist):
if update is not None:
print k
error[k] = np.max(
np.abs(te.hdf.root.clda[k]['filt_C'] - update['filt.C']))
print np.max(np.abs(error))
#!/usr/bin/python
'''
Test that the content of the npz file is the same as what is stored in the HDF file
'''
from db import dbfunctions as dbfn
from db.tracker import models
from tasks import performance
import plotutil
import tables
import numpy as np
te = performance._get_te(2440)
print te
te.clda_param_hist;
error = np.zeros(len(te.clda_param_hist))
for k, update in enumerate(te.clda_param_hist):
if update is not None:
print k
error[k] = np.max(np.abs(te.hdf.root.clda[k]['filt_C'] - update['filt.C']))
print np.max(np.abs(error))
#!/usr/bin/python
from db import dbfunctions as dbfn
reload(dbfn)
from tasks import bmimultitasks, performance
import numpy as np
from optparse import OptionParser
parser = OptionParser()
parser.add_option("-i", "--idx", dest="idx", type="int", help="block index to run verification", default=2298)
(options, args) = parser.parse_args()
idx = options.idx
te = performance._get_te(idx)
print te
hdf = te.hdf
dec = te.decoder
bmi_params = te.clda_param_hist
assist_level = hdf.root.task[:]['assist_level'].ravel()
spike_counts = hdf.root.task[:]['spike_counts']
target = hdf.root.task[:]['target']
cursor = hdf.root.task[:]['cursor']
task_msgs = hdf.root.task_msgs[:]
update_bmi_msgs = filter(lambda x: x['msg'] == 'update_bmi', task_msgs)
inds = [x[1] for x in update_bmi_msgs]
T = spike_counts.shape[0]
error = np.zeros(T)
from tasks import performance import matplotlib.pyplot as plt from riglib.bmi import robot_arms, assist, train, goal_calculators from riglib.stereo_opengl import ik import time from riglib import mp_calc reload(mp_calc) reload(assist) reload(robot_arms) reload(goal_calculators) pi = np.pi te = performance._get_te(3086) target = te.hdf.root.task[:]['target'] starting_pos_ws = np.array([5., 0, 5]) shoulder_anchor = np.array([2, 0, -15]) # Generate the list of targets from tasks import generatorfunctions target_list = generatorfunctions.centerout_2D_discrete(nblocks=1, ntargets=2) target_list = target_list.reshape(-1, 3) n_targets = target_list.shape[0] # Initialize the kinematic chain chain = robot_arms.PlanarXZKinematicChain([15, 15, 5, 5]) chain.joint_limits = [(-pi, pi), (-pi, 0), (-pi/2, pi/2), (-pi/2, 10*pi/180)]