def bin_array(y, nr_bins):
"""Takes circular array and digitises into n bins"""
y = np.array(y)
bins = np.arange(-math.pi - 0.000001, math.pi - (2 * math.pi / nr_bins),
(2 * math.pi / nr_bins))
y_bin = np.digitize(pycircstat.cdiff(y, 0), bins)
return y, y_bin
def calculate_bias(signed_error, target, other, subid, confidence = None):
targori = np.radians(target)
otherori = np.radians(other)
angdiff = circstat.cdiff(otherori, targori)
#demean error to remove any angular biases in overall responding (agnostic to any other data)
signed_error = np.subtract(signed_error, signed_error.mean())
nbin = 64 #64 bins
pbin = 0.25 #1/4 fo the data per bin
bins = circ_bini(angdiff, nbin, pbin)
biases = np.full(shape = nbin, fill_value = np.nan)
bincentre = np.full(shape = nbin, fill_value = np.nan)
precs = np.full(shape = nbin, fill_value = np.nan)
binid = np.full(shape = nbin, fill_value = 0)
for i in range(nbin): #loop over bins
biases[i] = np.mean(signed_error[bins[i,:]]) #sp.stats.circmean(signed_error[bins[i,:]], high = np.pi, low = -np.pi)
bincentre[i] = sp.stats.circmean(angdiff[bins[i,:]], high = np.pi, low = -np.pi)
precs[i] = sp.stats.circstd(signed_error[bins[i,:]], high = np.pi, low = -np.pi)
binid[i] = i
df = pd.DataFrame()
df['subid'] = np.full(shape = nbin, fill_value = subid)
df['bincentre'] = bincentre
df['bias'] = biases
df['prec'] = precs
df['binid'] = binid
return df
def compute_hilbert_for_cluster(self, this_cluster_name):
# first, get the eeg for just channels in cluster
cluster_rows = self.res['clusters'][this_cluster_name].notna()
cluster_elec_labels = self.res['clusters'][cluster_rows]['label']
cluster_eeg = self.subject_data[:, np.in1d(self.subject_data.channel, cluster_elec_labels)]
# bandpass eeg at the mean frequency, making sure the lower frequency isn't too low
cluster_mean_freq = self.res['clusters'][cluster_rows][this_cluster_name].mean()
cluster_freq_range = [cluster_mean_freq - self.hilbert_half_range, cluster_mean_freq + self.hilbert_half_range]
if cluster_freq_range[0] < SubjectTravelingWaveAnalysis.LOWER_MIN_FREQ:
cluster_freq_range[0] = SubjectTravelingWaveAnalysis.LOWER_MIN_FREQ
filtered_eeg = RAM_helpers.band_pass_eeg(cluster_eeg, cluster_freq_range)
filtered_eeg = filtered_eeg.transpose('channel', 'event', 'time')
# run the hilbert transform
complex_hilbert_res = hilbert(filtered_eeg.data, N=filtered_eeg.shape[-1], axis=-1)
# compute the phase of the filtered eeg
phase_data = filtered_eeg.copy()
phase_data.data = np.unwrap(np.angle(complex_hilbert_res))
# compute the power
power_data = filtered_eeg.copy()
power_data.data = np.abs(complex_hilbert_res) ** 2
# compute mean phase and phase difference between ref phase and each electrode phase
ref_phase = pycircstat.mean(phase_data.data, axis=0)
phase_data.data = pycircstat.cdiff(phase_data.data, ref_phase)
return phase_data, power_data, cluster_mean_freq
def plot_marginal_slices(ax, xx, yy, zz, selected_x, slice_yrange, slicegap):
"""
Marginal slices at different x values, as a function of y-axis
Parameters
----------
ax : axes object
xx : ndarray
yy : ndarray
zz : ndarray
selected_x : tuple
slice_yrange : tuple
slicegap : float
Returns
-------
"""
# Slices
yax = yy[:, 0] # y
xax = xx[0, :] # x
color_list = []
for idx, xval in enumerate(selected_x):
selected_xid = np.argmin(np.abs(xval - xax))
yrange_ids = [
np.argmin(np.abs(yax - y_bound_val))
for y_bound_val in slice_yrange
]
slice_cut = zz[yrange_ids[0]:yrange_ids[1], selected_xid]
slice_norm = slice_cut / np.sum(slice_cut)
slice_shifted = (slice_norm - slice_norm.min()) + idx * slicegap
yax_cut = yax[yrange_ids[0]:yrange_ids[1]]
slice_color = cm.brg(idx / selected_x.shape[0])
color_list.append(slice_color)
ax.plot(yax_cut, slice_shifted, c=slice_color)
y_cog = circmean(yax_cut, w=slice_norm)
slice_point = slice_shifted[np.argmin(np.abs(cdiff(yax_cut, y_cog)))]
marker, m_c = '.', 'k'
ax.scatter(y_cog, slice_point, marker=marker, c=[m_c], zorder=3.1)
ax.scatter(y_cog - 2 * np.pi,
slice_point,
marker=marker,
c=[m_c],
zorder=3.1)
ax.scatter(y_cog + 2 * np.pi,
slice_point,
marker=marker,
c=[m_c],
zorder=3.1)
return ax, color_list
def compute_hilbert_for_cluster(self, this_cluster_name):
# first, get the eeg for just channels in cluster
cluster_rows = self.res['clusters'][this_cluster_name].notna()
cluster_elec_labels = self.res['clusters'][cluster_rows]['label']
cluster_eeg = self.subject_data[:,
np.in1d(self.subject_data.
channel, cluster_elec_labels)]
# bandpass eeg at the mean frequency, making sure the lower frequency isn't too low
cluster_mean_freq = self.res['clusters'][cluster_rows][
this_cluster_name].mean()
cluster_freq_range = [
cluster_mean_freq - self.hilbert_half_range,
cluster_mean_freq + self.hilbert_half_range
]
if cluster_freq_range[0] < SubjectTravelingWaveAnalysis.LOWER_MIN_FREQ:
cluster_freq_range[0] = SubjectTravelingWaveAnalysis.LOWER_MIN_FREQ
filtered_eeg = ecog_helpers.band_pass_eeg(cluster_eeg,
cluster_freq_range)
filtered_eeg = filtered_eeg.transpose('channel', 'event', 'time')
# run the hilbert transform
complex_hilbert_res = hilbert(filtered_eeg.data,
N=filtered_eeg.shape[-1],
axis=-1)
# compute the phase of the filtered eeg
phase_data = filtered_eeg.copy()
phase_data.data = np.unwrap(np.angle(complex_hilbert_res))
# compute the power
power_data = filtered_eeg.copy()
power_data.data = np.abs(complex_hilbert_res)**2
# compute mean phase and phase difference between ref phase and each electrode phase
ref_phase = pycircstat.mean(phase_data.data, axis=0)
phase_data.data = pycircstat.cdiff(phase_data.data, ref_phase)
return phase_data, power_data, cluster_mean_freq
def test_circular_distance():
a = np.array([4.85065953, 0.79063862, 1.35698570])
assert_allclose(pycircstat.cdiff(a, a), np.zeros_like(a))
def compute_asymmetry(data, y=None, reference=None, min_deg=30, max_deg=60):
"""Compute distribution of x by phase-bins in the Instantaneous Frequency.
Parameters
----------
data : ndarray
three-dimensional array of [trial repeats by classes by time points].
Correct class is in position n_classes/2
*alternatively 'data', 'y', and 'reference' can be specified
as a dictionary
y : ndarray
Input vector for presented stimulus in radians with the same
length as trials
reference : ndarray
Input vector for reference stimulus in radians with the same
length as trials
min_deg : float
min angular distance cut-off point for trials to include
max_deg : float
max angular distance cut-off point for trials to include
Returns
-------
shift : ndarray
array containing asymmetry score for every time point.
CW : ndarray
array containing evidence for CW angular difference trials
CCW : ndarray
array containing evidence for CCW angular difference trials
"""
# check input
if type(data) is dict:
reference = data['reference']
y = data['y']
data = data['single_trial_ev_centered']
if type(data) is np.ndarray:
if y is None or reference is None:
raise Exception("Specifiy both y and reference variable") # check
if y.max() > (2 * np.pi) or reference.max() > (2 * np.pi):
# if variables are out of range
raise Exception("Define y and reference in radians")
# main body
n_trials, n_classes, n_tps = data.shape
angular_difference = pycircstat.cdiff(reference, y)
# compute evidence for all classes for trials CW from current angle
CW = data[(angular_difference <= -min_deg / 90 * np.pi) &
(angular_difference > -max_deg / 90 * np.pi), :, :].mean(0)
# compute evidence for all classes for trials CCW from current angle
CCW = data[(angular_difference >= min_deg / 90 * np.pi) &
(angular_difference < max_deg / 90 * np.pi), :, :].mean(0)
# sum
shift = ((CW[0:int(n_classes / 2 - 1), :].mean(0) -
CW[int(n_classes / 2):(n_classes - 1), :].mean(0)) -
(CCW[0:int(n_classes / 2 - 1), :].mean(0) -
CCW[int(n_classes / 2):(n_classes - 1), :].mean(0)))
return shift, CW, CCW
def analysis(self):
"""
"""
if self.subject_data is None:
print('%s: compute or load data first with .load_data()!' % self.subject)
# Get recalled or not labels
if self.recall_filter_func is None:
print('%s SME: please provide a .recall_filter_func function.' % self.subject)
recalled = self.recall_filter_func(self.subject_data)
# filter to electrodes in ROIs. First get broad electrode region labels
region_df = self.bin_eloctrodes_into_rois()
region_df['merged_col'] = region_df['hemi'] + '-' + region_df['region']
# make sure we have electrodes in each unique region
for roi in self.roi_list:
has_elecs = []
for label in roi:
if np.any(region_df.merged_col == label):
has_elecs.append(True)
if ~np.any(has_elecs):
print('{}: no {} electrodes, cannot compute synchrony.'.format(self.subject, roi))
return
# then filter into just to ROIs defined above
elecs_to_use = region_df.merged_col.isin([item for sublist in self.roi_list for item in sublist])
elec_scheme = self.elec_info.copy(deep=True)
elec_scheme['ROI'] = region_df.merged_col[elecs_to_use]
elec_scheme = elec_scheme[elecs_to_use].reset_index()
if self.use_wavelets:
phase_data = MorletWaveletFilter(self.subject_data[:, elecs_to_use], self.wavelet_freq,
output='phase', width=5, cpus=12,
verbose=False).filter()
else:
# band pass eeg
phase_data = RAM_helpers.band_pass_eeg(self.subject_data[:, elecs_to_use], self.hilbert_band_pass_range)
# get phase at each timepoint
phase_data.data = np.angle(hilbert(phase_data.data, N=phase_data.shape[-1], axis=-1))
# remove the buffer
phase_data = phase_data.remove_buffer(self.buf_ms / 1000.)
# loop over each pair of ROIs
for region_pair in combinations(self.roi_list, 2):
elecs_region_1 = np.where(elec_scheme.ROI.isin(region_pair[0]))[0]
elecs_region_2 = np.where(elec_scheme.ROI.isin(region_pair[1]))[0]
elec_label_pairs = []
elec_pair_pvals = []
elec_pair_zs = []
elec_pair_rvls = []
elec_pair_pvals_rec = []
elec_pair_zs_rec = []
elec_pair_rvls_rec = []
elec_pair_pvals_nrec = []
elec_pair_zs_nrec = []
elec_pair_rvls_nrec = []
delta_mem_rayleigh_zscores = []
delta_mem_rvl_zscores = []
elec_pair_phase_diffs = []
# loop over all pairs of electrodes in the ROIs
for elec_1 in elecs_region_1:
for elec_2 in elecs_region_2:
elec_label_pairs.append([elec_scheme.iloc[elec_1].label, elec_scheme.iloc[elec_2].label])
# and take the difference in phase values for this electrode pair
elec_pair_phase_diff = pycircstat.cdiff(phase_data[:, elec_1], phase_data[:, elec_2])
if self.include_phase_diffs_in_res:
elec_pair_phase_diffs.append(elec_pair_phase_diff)
# compute the circular stats
elec_pair_stats = calc_circ_stats(elec_pair_phase_diff, recalled, do_perm=False)
elec_pair_pvals.append(elec_pair_stats['elec_pair_pval'])
elec_pair_zs.append(elec_pair_stats['elec_pair_z'])
elec_pair_rvls.append(elec_pair_stats['elec_pair_rvl'])
elec_pair_pvals_rec.append(elec_pair_stats['elec_pair_pval_rec'])
elec_pair_zs_rec.append(elec_pair_stats['elec_pair_z_rec'])
elec_pair_pvals_nrec.append(elec_pair_stats['elec_pair_pval_nrec'])
elec_pair_zs_nrec.append(elec_pair_stats['elec_pair_z_nrec'])
elec_pair_rvls_rec.append(elec_pair_stats['elec_pair_rvl_rec'])
elec_pair_rvls_nrec.append(elec_pair_stats['elec_pair_rvl_nrec'])
# compute null distributions for the memory stats
if self.do_perm_test:
delta_mem_rayleigh_zscore, delta_mem_rvl_zscore = self.compute_null_stats(elec_pair_phase_diff,
recalled,
elec_pair_stats)
delta_mem_rayleigh_zscores.append(delta_mem_rayleigh_zscore)
delta_mem_rvl_zscores.append(delta_mem_rvl_zscore)
region_pair_key = '+'.join(['-'.join(r) for r in region_pair])
self.res[region_pair_key] = {}
self.res[region_pair_key]['elec_label_pairs'] = elec_label_pairs
self.res[region_pair_key]['elec_pair_pvals'] = np.stack(elec_pair_pvals, 0)
self.res[region_pair_key]['elec_pair_zs'] = np.stack(elec_pair_zs, 0)
self.res[region_pair_key]['elec_pair_rvls'] = np.stack(elec_pair_rvls, 0)
self.res[region_pair_key]['elec_pair_pvals_rec'] = np.stack(elec_pair_pvals_rec, 0)
self.res[region_pair_key]['elec_pair_zs_rec'] = np.stack(elec_pair_zs_rec, 0)
self.res[region_pair_key]['elec_pair_pvals_nrec'] = np.stack(elec_pair_pvals_nrec, 0)
self.res[region_pair_key]['elec_pair_zs_nrec'] = np.stack(elec_pair_zs_nrec, 0)
self.res[region_pair_key]['elec_pair_rvls_rec'] = np.stack(elec_pair_rvls_rec, 0)
self.res[region_pair_key]['elec_pair_rvls_nrec'] = np.stack(elec_pair_rvls_nrec, 0)
if self.do_perm_test:
self.res[region_pair_key]['delta_mem_rayleigh_zscores'] = np.stack(delta_mem_rayleigh_zscores, 0)
self.res[region_pair_key]['delta_mem_rvl_zscores'] = np.stack(delta_mem_rvl_zscores, 0)
if self.include_phase_diffs_in_res:
self.res[region_pair_key]['elec_pair_phase_diffs'] = np.stack(elec_pair_phase_diffs, -1)
self.res[region_pair_key]['time'] = phase_data.time.data
self.res[region_pair_key]['recalled'] = recalled
def analysis(self):
"""
Runs the phase synchrony analysis.
"""
if self.subject_data is None:
print('%s: compute or load data first with .load_data()!' % self.subject)
# Get recalled or not labels
if self.recall_filter_func is None:
print('%s SME: please provide a .recall_filter_func function.' % self.subject)
recalled = self.recall_filter_func(self.subject_data)
# filter to electrodes in ROIs. First get broad electrode region labels
region_df = self.bin_eloctrodes_into_rois()
region_df['merged_col'] = region_df['hemi'] + '-' + region_df['region']
# make sure we have electrodes in each unique region
for roi in self.roi_list:
has_elecs = []
for label in roi:
if np.any(region_df.merged_col == label):
has_elecs.append(True)
if ~np.any(has_elecs):
print('{}: no {} electrodes, cannot compute synchrony.'.format(self.subject, roi))
return
# then filter into just to ROIs defined above
elecs_to_use = region_df.merged_col.isin([item for sublist in self.roi_list for item in sublist])
elec_scheme = self.elec_info.copy(deep=True)
elec_scheme['ROI'] = region_df.merged_col[elecs_to_use]
elec_scheme = elec_scheme[elecs_to_use].reset_index()
if self.use_wavelets:
phase_data = MorletWaveletFilter(self.subject_data[:, elecs_to_use], self.wavelet_freq,
output='phase', width=5, cpus=12,
verbose=False).filter()
else:
# band pass eeg
phase_data = ecog_helpers.band_pass_eeg(self.subject_data[:, elecs_to_use], self.hilbert_band_pass_range)
# get phase at each timepoint
phase_data.data = np.angle(hilbert(phase_data.data, N=phase_data.shape[-1], axis=-1))
# remove the buffer
phase_data = phase_data.remove_buffer(self.buf_ms / 1000.)
# loop over each pair of ROIs
for region_pair in combinations(self.roi_list, 2):
elecs_region_1 = np.where(elec_scheme.ROI.isin(region_pair[0]))[0]
elecs_region_2 = np.where(elec_scheme.ROI.isin(region_pair[1]))[0]
elec_label_pairs = []
elec_pair_pvals = []
elec_pair_zs = []
elec_pair_rvls = []
elec_pair_pvals_rec = []
elec_pair_zs_rec = []
elec_pair_rvls_rec = []
elec_pair_pvals_nrec = []
elec_pair_zs_nrec = []
elec_pair_rvls_nrec = []
delta_mem_rayleigh_zscores = []
delta_mem_rvl_zscores = []
elec_pair_phase_diffs = []
# loop over all pairs of electrodes in the ROIs
for elec_1 in elecs_region_1:
for elec_2 in elecs_region_2:
elec_label_pairs.append([elec_scheme.iloc[elec_1].label, elec_scheme.iloc[elec_2].label])
# and take the difference in phase values for this electrode pair
elec_pair_phase_diff = pycircstat.cdiff(phase_data[:, elec_1], phase_data[:, elec_2])
if self.include_phase_diffs_in_res:
elec_pair_phase_diffs.append(elec_pair_phase_diff)
# compute the circular stats
elec_pair_stats = calc_circ_stats(elec_pair_phase_diff, recalled, do_perm=False)
elec_pair_pvals.append(elec_pair_stats['elec_pair_pval'])
elec_pair_zs.append(elec_pair_stats['elec_pair_z'])
elec_pair_rvls.append(elec_pair_stats['elec_pair_rvl'])
elec_pair_pvals_rec.append(elec_pair_stats['elec_pair_pval_rec'])
elec_pair_zs_rec.append(elec_pair_stats['elec_pair_z_rec'])
elec_pair_pvals_nrec.append(elec_pair_stats['elec_pair_pval_nrec'])
elec_pair_zs_nrec.append(elec_pair_stats['elec_pair_z_nrec'])
elec_pair_rvls_rec.append(elec_pair_stats['elec_pair_rvl_rec'])
elec_pair_rvls_nrec.append(elec_pair_stats['elec_pair_rvl_nrec'])
# compute null distributions for the memory stats
if self.do_perm_test:
delta_mem_rayleigh_zscore, delta_mem_rvl_zscore = self.compute_null_stats(elec_pair_phase_diff,
recalled,
elec_pair_stats)
delta_mem_rayleigh_zscores.append(delta_mem_rayleigh_zscore)
delta_mem_rvl_zscores.append(delta_mem_rvl_zscore)
region_pair_key = '+'.join(['-'.join(r) for r in region_pair])
self.res[region_pair_key] = {}
self.res[region_pair_key]['elec_label_pairs'] = elec_label_pairs
self.res[region_pair_key]['elec_pair_pvals'] = np.stack(elec_pair_pvals, 0)
self.res[region_pair_key]['elec_pair_zs'] = np.stack(elec_pair_zs, 0)
self.res[region_pair_key]['elec_pair_rvls'] = np.stack(elec_pair_rvls, 0)
self.res[region_pair_key]['elec_pair_pvals_rec'] = np.stack(elec_pair_pvals_rec, 0)
self.res[region_pair_key]['elec_pair_zs_rec'] = np.stack(elec_pair_zs_rec, 0)
self.res[region_pair_key]['elec_pair_pvals_nrec'] = np.stack(elec_pair_pvals_nrec, 0)
self.res[region_pair_key]['elec_pair_zs_nrec'] = np.stack(elec_pair_zs_nrec, 0)
self.res[region_pair_key]['elec_pair_rvls_rec'] = np.stack(elec_pair_rvls_rec, 0)
self.res[region_pair_key]['elec_pair_rvls_nrec'] = np.stack(elec_pair_rvls_nrec, 0)
if self.do_perm_test:
self.res[region_pair_key]['delta_mem_rayleigh_zscores'] = np.stack(delta_mem_rayleigh_zscores, 0)
self.res[region_pair_key]['delta_mem_rvl_zscores'] = np.stack(delta_mem_rvl_zscores, 0)
if self.include_phase_diffs_in_res:
self.res[region_pair_key]['elec_pair_phase_diffs'] = np.stack(elec_pair_phase_diffs, -1)
self.res[region_pair_key]['time'] = phase_data.time.data
self.res[region_pair_key]['recalled'] = recalled
def test_circular_distance():
a = np.array([4.85065953, 0.79063862, 1.35698570])
assert_allclose(pycircstat.cdiff(a, a), np.zeros_like(a))
def calculate_bias_covar(signed_error, target, other, subid, nbin = 64, pbin = 0.25, covariate = None, covariate_label = None, median_split = None):
targori = np.radians(target)
otherori = np.radians(other)
angdiff = circstat.cdiff(otherori, targori)
#demean error to remove any angular biases in overall responding (agnostic to any other data)
signed_error = np.subtract(signed_error, signed_error.mean())
# #params for creating data bins
# nbin = 64 #64 bins
# pbin = 0.25 #1/4 fo the data per bin
if median_split == True and covariate is not None: #check there's a covar of interest and that we need to median split data for it
covar_median = np.median(covariate)
belowmed_trls = np.less_equal(covariate, covar_median)
abovemed_trls = np.greater(covariate, covar_median)
belowmed_vars = dict()
abovemed_vars = dict()
belowmed_vars['angdiff'] = angdiff[belowmed_trls]
belowmed_vars['targori'] = targori[belowmed_trls]
belowmed_vars['otherori'] = otherori[belowmed_trls]
belowmed_vars['signed_error'] = signed_error[belowmed_trls]
belowmed_vars[covariate_label] = covariate[belowmed_trls]
abovemed_vars['angdiff'] = angdiff[abovemed_trls]
abovemed_vars['targori'] = targori[abovemed_trls]
abovemed_vars['otherori'] = otherori[abovemed_trls]
abovemed_vars['signed_error'] = signed_error[abovemed_trls]
abovemed_vars[covariate_label] = covariate[abovemed_trls]
belowmed_bins = circ_bini(belowmed_vars['angdiff'], nbin, pbin)
abovemed_bins = circ_bini(abovemed_vars['angdiff'], nbin, pbin)
below_vars = dict()
above_vars = dict()
for key in ['biases', 'bincentre', 'precs', 'binid']:
below_vars[key] = np.full(shape = nbin, fill_value = np.nan)
above_vars[key] = np.full(shape = nbin, fill_value = np.nan)
for i in range(nbin): #loop over bins
below_vars['biases'][i] = np.mean(belowmed_vars['signed_error'][belowmed_bins[i,:]])
below_vars['bincentre'][i] = sp.stats.circmean( belowmed_vars['angdiff'][belowmed_bins[i,:]], high = np.pi, low = -np.pi )
below_vars['precs'][i] = sp.stats.circstd( belowmed_vars['signed_error'][belowmed_bins[i,:]], high = np.pi, low = -np.pi )
below_vars['binid'][i] = i
above_vars['biases'][i] = np.mean(abovemed_vars['signed_error'][abovemed_bins[i,:]])
above_vars['bincentre'][i] = sp.stats.circmean( abovemed_vars['angdiff'][abovemed_bins[i,:]], high = np.pi, low = -np.pi )
above_vars['precs'][i] = sp.stats.circstd( abovemed_vars['signed_error'][abovemed_bins[i,:]], high = np.pi, low = -np.pi )
above_vars['binid'][i] = i
above_df = pd.DataFrame(above_vars); above_df['median_split'] = 'above'; above_df.binid = above_df.binid.astype(int)
below_df = pd.DataFrame(below_vars); below_df['median_split'] = 'below'; below_df.binid = below_df.binid.astype(int)
df = pd.concat([above_df, below_df])
else:
bins = circ_bini(angdiff, nbin, pbin)
biases = np.full(shape = nbin, fill_value = np.nan)
bincentre = np.full(shape = nbin, fill_value = np.nan)
precs = np.full(shape = nbin, fill_value = np.nan)
binid = np.full(shape = nbin, fill_value = 0)
for i in range(nbin): #loop over bins
biases[i] = np.mean(signed_error[bins[i,:]]) #sp.stats.circmean(signed_error[bins[i,:]], high = np.pi, low = -np.pi)
bincentre[i] = sp.stats.circmean(angdiff[bins[i,:]], high = np.pi, low = -np.pi)
precs[i] = sp.stats.circstd(signed_error[bins[i,:]], high = np.pi, low = -np.pi)
binid[i] = i
df = pd.DataFrame()
df['subid'] = np.full(shape = nbin, fill_value = subid)
df['bincentre'] = bincentre
df['bias'] = biases
df['prec'] = precs
df['binid'] = binid
df['subid'] = subid
return df
# #fitting evidence
# classifier_output[:,sb_count,2] = evidence['cos_convolved']
# convolve with cosine and obtain evidence
evidence['single_trial_evidence'] = evidence['single_trial_evidence_store']
y, evidence['y'] = bin_array(np.array(df_read['presented'])[inx],nr_bins)
evidence = cos_convolve(evidence)
# #fitting evidence
classifier_output[:,sb_count,0] = evidence['cos_convolved']
classifier_output_tuning[:,:,sb_count,0] = evidence['centered_prediction']
# df_read.to_csv((projectloc + '/saved_data_forNick/behavioural_betweentrial_S%02d_June2021.csv' %sb_count))
# scipy.io.savemat(projectloc + '/saved_data_forNick/LDA_betweentrial_S%02d_June2021.mat' %sb_count, {'data':evidence['single_trial_ev_centered']})
df_read['diff_stim_prev_pang'] = pycircstat.cdiff(df_read['prev_non_probe_ang'],df_read['presented'])
min_deg = 0 #2.903#2.8125 #10
max_deg = 60 #58.06 #56.25 #50
# right[:,:,sb_count] = evidence['single_trial_ev_centered'][(df_read['diff_stim_prev_pang'][inx] >= min_deg/90*np.pi) & (df_read['diff_stim_prev_pang'][inx] < max_deg/90*np.pi),:,:].mean(0)
right[:,:,sb_count] = evidence['single_trial_ev_centered'][inx2][(df_read['diff_stim_prev_pang'][inx][inx2] >= min_deg/90*np.pi) & (df_read['diff_stim_prev_pang'][inx][inx2] < max_deg/90*np.pi),:,:].mean(0)
# left[:,:,sb_count] = evidence['single_trial_ev_centered'][(df_read['diff_stim_prev_pang'][inx] <= -min_deg/90*np.pi) & (df_read['diff_stim_prev_pang'][inx] > -max_deg/90*np.pi),:,:].mean(0)
left[:,:,sb_count] = evidence['single_trial_ev_centered'][inx2][(df_read['diff_stim_prev_pang'][inx][inx2] <= -min_deg/90*np.pi) & (df_read['diff_stim_prev_pang'][inx][inx2] > -max_deg/90*np.pi),:,:].mean(0)
shift[:,sb_count] = (left[0:4,:,sb_count].mean(0) - left[5:9,:,sb_count].mean(0)) - (right[0:4,:,sb_count].mean(0) - right[5:9,:,sb_count].mean(0))
#%%
# stim_prev = {}
# stim_prev['Crossdec_prev_probe_Protect'] = classifier_output[:,:,2]
# stim_prev['shift_prev_probe_Protect'] = shift
