def _plvstat(xp1, xp2, plv, n_perm, n_jobs, perm):
"""Sub plv-stat function
"""
# Compute plv for each permutation of xp2 trials :
plvs = np.array(
Parallel(n_jobs=n_jobs)(delayed(_plv)(xp1, xp2[..., p]) for p in perm))
# Get p-values from permutations :
return perm_2pvalue(plv, plvs, n_perm, tail=1)
def _plvstat(xp1, xp2, plv, n_perm, n_jobs, perm):
"""Sub plv-stat function
"""
# Compute plv for each permutation of xp2 trials :
plvs = np.array(Parallel(n_jobs=n_jobs)(delayed(_plv)(
xp1, xp2[..., p]) for p in perm))
# Get p-values from permutations :
return perm_2pvalue(plv, plvs, n_perm, tail=1)
def fit(self, x, mf=False, center=False, grp=None,
method='bino', n_perm=200, rndstate=0, n_jobs=-1):
"""Apply the classification and cross-validation objects to the array x.
Args:
x: list
List of dataset for each subject. All the dataset in the list
should have the same number of columns but the number of lines
could be diffrent for each subject and must correspond to the
same number of lines each each label vector of y.
Kargs:
mf: bool, optional, [def: False]
If mf=False, the returned decoding accuracy (da) will have a
shape of (1, rep) where rep, is the number of repetitions.
This mean that all the features are used together. If mf=True,
da.shape = (M, rep), where M is the number of columns of x.
center: optional, bool, [def: False]
Normalize fatures with a zero mean by substracting then dividing
by the mean. The center parameter should be set to True if the
classifier is a svm.
grp: array, optional, [def: None]
If mf=True, the grp parameter allow to define group of features.
If x.shape = (N, 5) and grp=np.array([0,0,1,2,1]), this mean that
3 groups of features will be considered : (0,1,2)
method: string, optional, [def: 'bino']
Four methods are implemented to test the statistical significiance
of the decoding accuracy :
- 'bino': binomial test
- 'label_rnd': randomly shuffle the labels
Methods 2 and 3 are based on permutations. They should provide
similar results. But 4 should be more conservative.
n_perm: integer, optional, [def: 200]
Number of permutations for the methods 2, 3 and 4
rndstate: integer, optional, [def: 0]
Fix the random state of the machine. Usefull to reproduce results.
n_jobs: integer, optional, [def: -1]
Control the number of jobs to cumpute the decoding accuracy. If
n_jobs = -1, all the jobs are used.
Return:
da: array
The decoding accuracy of shape n_repetitions x n_features
pvalue: array
Array of associated pvalue of shape n_features
daPerm: array
Array of all the decodings obtained for each permutations of shape
n_perm x n_features
.. rubric:: Footnotes
.. [#f8] `Ojala and Garriga, 2010 <http://www.jmlr.org/papers/volume11/ojala10a/ojala10a.pdf>`_
.. [#f9] `Combrisson and Jerbi, 2015 <http://www.ncbi.nlm.nih.gov/pubmed/25596422/>`_
"""
# Check x, y:
xbk = x.copy()
x, y, train, test = _checkXY(x, self._y, mf, grp, center, self)
nsuj, nfeat = x.shape
# Run classification:
da, ytrue, ypred = _fit(x, y, train, test, self, n_jobs)
# Get statistics:
# -------------------------------------------------------------
# Binomial :
# -------------------------------------------------------------
if method == 'bino':
pvalue = bino_da2p(self._ry, da)
daperm = None
pperm = None
# -------------------------------------------------------------
# Permutations :
# -------------------------------------------------------------
# -> Shuffle the labels :
elif method == 'label_rnd':
y_sh = [_checkXY(xbk, [np.random.permutation(i) for i in self._y],
mf, grp, center, self)[1] for k in range(n_perm)]
cvs = Parallel(n_jobs=n_jobs)(delayed(_fit)(
x, y_sh[k], train, test, self, 1)
for k in range(n_perm))
# Reconstruct daperm and get the associated p-value:
daperm, _, _ = zip(*cvs)
daperm = np.array(daperm).reshape(n_perm, nfeat)
pvalue = perm_2pvalue(da, daperm, n_perm, tail=1)
pperm = pvalue
else:
raise ValueError('No statistical method '+method+' found')
# Try to get featinfo:
try:
if grp is not None:
grp = uorderlst(grp)
else:
grp = np.arange(nfeat)
self.info.featinfo = self.info._featinfo(self._clf, self._cv,
da[:, np.newaxis], grp=grp,
pperm=pperm)
except:
pass
return da, pvalue, daperm
def fit(self,
x,
mf=False,
center=False,
grp=None,
method='bino',
n_perm=200,
rndstate=0,
n_jobs=-1):
"""Apply the classification and cross-validation objects to the array x.
Args:
x: array
Data to classify. Consider that x.shape = (N, M), N is the number
of trials (which should be the length of y). M, the number of
colums, is a supplementar dimension for classifying data. If M = 1,
the data is consider as a single feature. If M > 1, use the
parameter mf to say if x should be consider as a single feature
(mf=False) or multi-features (mf=True)
Kargs:
mf: bool, optional, [def: False]
If mf=False, the returned decoding accuracy (da) will have a
shape of (1, rep) where rep, is the number of repetitions.
This mean that all the features are used together. If mf=True,
da.shape = (M, rep), where M is the number of columns of x.
center: optional, bool, [def: False]
Normalize fatures with a zero mean by substracting then dividing
by the mean. The center parameter should be set to True if the
classifier is a svm.
grp: array, optional, [def: None]
If mf=True, the grp parameter allow to define group of features.
If x.shape = (N, 5) and grp=np.array([0,0,1,2,1]), this mean that
3 groups of features will be considered : (0,1,2)
method: string, optional, [def: 'bino']
Four methods are implemented to test the statistical significiance
of the decoding accuracy :
- 'bino': binomial test
- 'label_rnd': randomly shuffle the labels
- 'full_rnd': randomly shuffle the whole array x
- 'intra_rnd': randomly shuffle x inside each class and each feature
Methods 2, 3 and 4 are based on permutations. The method 2 and 3
should provide similar results. But 4 should be more conservative.
n_perm: integer, optional, [def: 200]
Number of permutations for the methods 2, 3 and 4
rndstate: integer, optional, [def: 0]
Fix the random state of the machine. Usefull to reproduce results.
n_jobs: integer, optional, [def: -1]
Control the number of jobs to cumpute the decoding accuracy. If
n_jobs = -1, all the jobs are used.
Return:
da: array
The decoding accuracy of shape n_repetitions x n_features
pvalue: array
Array of associated pvalue of shape n_features
daPerm: array
Array of all the decodings obtained for each permutations of shape
n_perm x n_features
.. rubric:: Footnotes
.. [#f8] `Ojala and Garriga, 2010 <http://www.jmlr.org/papers/volume11/ojala10a/ojala10a.pdf>`_
.. [#f9] `Combrisson and Jerbi, 2015 <http://www.ncbi.nlm.nih.gov/pubmed/25596422/>`_
"""
# Get the true decoding accuracy:
da, x, y, self._ytrue, self._ypred = _fit(x, self._y, self._clf,
self._cv.cvr, mf, grp,
center, n_jobs)
nfeat = len(x)
rndstate = np.random.RandomState(rndstate)
score = np.array([np.mean(k) for k in da])
# Get statistics:
# -------------------------------------------------------------
# Binomial :
# -------------------------------------------------------------
if method == 'bino':
pvalue = bino_da2p(y, score)
daPerm = None
pperm = None
# -------------------------------------------------------------
# Permutations :
# -------------------------------------------------------------
elif method.lower().find('_rnd') + 1:
# Generate idx tricks :
iteract = product(range(n_perm), range(nfeat))
# -> Shuffle the labels :
if method == 'label_rnd':
y_sh = [rndstate.permutation(y) for k in range(n_perm)]
cvs = Parallel(n_jobs=n_jobs)(delayed(_cvscore)(
x[k], y_sh[i], clone(self._clf), self._cv.cvr[0])
for i, k in iteract)
# -> Full randomization :
elif method == 'full_rnd':
cvs = Parallel(n_jobs=n_jobs)(
delayed(_cvscore)(rndstate.permutation(x[k]), y,
clone(self._clf), self._cv.cvr[0])
for i, k in iteract)
# -> Shuffle intra-class :
elif method == 'intra_rnd':
cvs = Parallel(n_jobs=n_jobs)(
delayed(_cvscore)(x[k][permIntraClass(y, rnd=i), :], y,
clone(self._clf), self._cv.cvr[0])
for i, k in iteract)
# Reconstruct daPerm and get the associated p-value:
daPerm, _, _ = zip(*cvs)
daPerm = np.array(daPerm).reshape(n_perm, nfeat)
pvalue = perm_2pvalue(score, daPerm, n_perm, tail=1)
pperm = pvalue
else:
raise ValueError('No statistical method ' + method + ' found')
# Get features informations:
try:
if grp is not None:
grp = uorderlst(grp)
self.info.featinfo = self.info._featinfo(self._clf,
self._cv,
da,
grp=grp,
pperm=pperm)
except:
pass
return da.T, pvalue, daPerm
def fit(self, x, mf=False, center=False, grp=None,
method='bino', n_perm=200, rndstate=0, n_jobs=-1):
"""Apply the classification and cross-validation objects to the array x.
Args:
x: array
Data to classify. Consider that x.shape = (N, M), N is the number
of trials (which should be the length of y). M, the number of
colums, is a supplementar dimension for classifying data. If M = 1,
the data is consider as a single feature. If M > 1, use the
parameter mf to say if x should be consider as a single feature
(mf=False) or multi-features (mf=True)
Kargs:
mf: bool, optional, [def: False]
If mf=False, the returned decoding accuracy (da) will have a
shape of (1, rep) where rep, is the number of repetitions.
This mean that all the features are used together. If mf=True,
da.shape = (M, rep), where M is the number of columns of x.
center: optional, bool, [def: False]
Normalize fatures with a zero mean by substracting then dividing
by the mean. The center parameter should be set to True if the
classifier is a svm.
grp: array, optional, [def: None]
If mf=True, the grp parameter allow to define group of features.
If x.shape = (N, 5) and grp=np.array([0,0,1,2,1]), this mean that
3 groups of features will be considered : (0,1,2)
method: string, optional, [def: 'bino']
Four methods are implemented to test the statistical significiance
of the decoding accuracy :
- 'bino': binomial test
- 'label_rnd': randomly shuffle the labels
- 'full_rnd': randomly shuffle the whole array x
- 'intra_rnd': randomly shuffle x inside each class and each feature
Methods 2, 3 and 4 are based on permutations. The method 2 and 3
should provide similar results. But 4 should be more conservative.
n_perm: integer, optional, [def: 200]
Number of permutations for the methods 2, 3 and 4
rndstate: integer, optional, [def: 0]
Fix the random state of the machine. Usefull to reproduce results.
n_jobs: integer, optional, [def: -1]
Control the number of jobs to cumpute the decoding accuracy. If
n_jobs = -1, all the jobs are used.
Return:
da: array
The decoding accuracy of shape n_repetitions x n_features
pvalue: array
Array of associated pvalue of shape n_features
daPerm: array
Array of all the decodings obtained for each permutations of shape
n_perm x n_features
.. rubric:: Footnotes
.. [#f8] `Ojala and Garriga, 2010 <http://www.jmlr.org/papers/volume11/ojala10a/ojala10a.pdf>`_
.. [#f9] `Combrisson and Jerbi, 2015 <http://www.ncbi.nlm.nih.gov/pubmed/25596422/>`_
"""
# Get the true decoding accuracy:
da, x, y, self._ytrue, self._ypred = _fit(x, self._y, self._clf, self._cv.cvr,
mf, grp, center, n_jobs)
nfeat = len(x)
rndstate = np.random.RandomState(rndstate)
score = np.array([np.mean(k) for k in da])
# Get statistics:
# -------------------------------------------------------------
# Binomial :
# -------------------------------------------------------------
if method == 'bino':
pvalue = bino_da2p(y, score)
daPerm = None
pperm = None
# -------------------------------------------------------------
# Permutations :
# -------------------------------------------------------------
elif method.lower().find('_rnd')+1:
# Generate idx tricks :
iteract = product(range(n_perm), range(nfeat))
# -> Shuffle the labels :
if method == 'label_rnd':
y_sh = [rndstate.permutation(y) for k in range(n_perm)]
cvs = Parallel(n_jobs=n_jobs)(delayed(_cvscore)(
x[k], y_sh[i], clone(self._clf), self._cv.cvr[0])
for i, k in iteract)
# -> Full randomization :
elif method == 'full_rnd':
cvs = Parallel(n_jobs=n_jobs)(delayed(_cvscore)(
rndstate.permutation(x[k]), y, clone(self._clf),
self._cv.cvr[0]) for i, k in iteract)
# -> Shuffle intra-class :
elif method == 'intra_rnd':
cvs = Parallel(n_jobs=n_jobs)(delayed(_cvscore)(
x[k][permIntraClass(y, rnd=i), :], y, clone(self._clf),
self._cv.cvr[0]) for i, k in iteract)
# Reconstruct daPerm and get the associated p-value:
daPerm, _, _ = zip(*cvs)
daPerm = np.array(daPerm).reshape(n_perm, nfeat)
pvalue = perm_2pvalue(score, daPerm, n_perm, tail=1)
pperm = pvalue
else:
raise ValueError('No statistical method '+method+' found')
# Get features informations:
try:
if grp is not None:
grp = uorderlst(grp)
self.info.featinfo = self.info._featinfo(self._clf, self._cv, da,
grp=grp, pperm=pperm)
except:
pass
return da.T, pvalue, daPerm
def get(self, xpha, xamp, n_perm=200, p=0.05, matricial=False, n_jobs=-1):
"""Get the normalized cfc mesure between an xpha and xamp signals.
Args:
xpha: array
Signal for phase. The shape of xpha should be :
(n_electrodes x n_pts x n_trials)
xamp: array
Signal for amplitude. The shape of xamp should be :
(n_electrodes x n_pts x n_trials)
Kargs:
n_perm: integer, optional, [def: 200]
Number of permutations for normalizing the cfc mesure.
p: float, optional, [def: 0.05]
p-value for the statistical method of Ozkurt 2012.
matricial: bool, optional, [def: False]
Some methods can work in matricial computation. This can lead
to a 10x or 30x time faster. But, please, monitor your RAM usage
beacause this parameter can use a lot of RAM. So, turn this parameter
in case of small computation.
n_jobs: integer, optional, [def: -1]
Control the number of jobs for parallel computing. Use 1, 2, ..
depending of your number or cores. -1 for all the cores.
If the same signal is used (example : xpha=x and xamp=x), this mean
the program compute a local cfc.
Returns:
ncfc: array
The cfc mesure of size :
(n_amplitude x n_phase x n_electrodes x n_windows x n_trials)
pvalue: array
The associated p-values of size :
(n_amplitude x n_phase x n_electrodes x n_windows)
"""
# Check the inputs variables :
xpha, xamp = _cfcCheck(xpha, xamp, self._npts)
self.n_perm = n_perm
self._matricial = matricial
if n_perm != 0:
self.p = 1 / n_perm
else:
self.p = None
N = xpha.shape[0]
# Manage jobs repartition :
if (N < cpu_count()) and (n_jobs != 1):
surJob = n_jobs
elecJob = 1
elif (N >= cpu_count()) and (n_jobs != 1):
surJob = 1
elecJob = n_jobs
else:
surJob, elecJob = 1, 1
# Get the unormalized cfc and surogates:
cfcsu = Parallel(n_jobs=elecJob)(
delayed(_cfcFiltSuro)(xpha[k, ...], xamp[k, ...], surJob, self)
for k in range(N))
uCfc, Suro, mSuro, stdSuro = zip(*cfcsu)
uCfc = np.array(uCfc)
# Permutations ans stat:
if (self.Id[0] is not '5'):
# Compute permutations :
if (self.n_perm is not 0) and (self.Id[1] is not '0'):
Suro, mSuro, stdSuro = np.array(Suro), np.array(
mSuro), np.array(stdSuro)
# Normalize each cfc:
_, _, Norm, _, _, _ = CfcSettings(self.Id)
nCfc = Norm(uCfc, mSuro, stdSuro)
# Confidence interval :
pvalue = perm_2pvalue(uCfc.mean(2),
np.rollaxis(Suro.mean(2), 4),
self.n_perm,
tail=1)
return nCfc.transpose(3, 4, 0, 1,
2), pvalue.transpose(2, 3, 0, 1)
else:
return uCfc.transpose(3, 4, 0, 1, 2), None
elif self.Id[0] is '5':
# Ozkurt threshold :
xlim = (erfinv(1 - p)**2)
# Set to zero non-significant values:
idxUn = np.where(uCfc <= 2 * xlim)
uCfc[idxUn] = 0
return uCfc.transpose(3, 4, 0, 1, 2), None
def _evalstat(self, x, bsl):
"""Statistical evaluation of features
[x] = [xn] = (nFce, npts, nTrials)
"""
# Unpack variables:
statmeth = self._statmeth
n_perm = self._n_perm
tail = self._2t
maxst = self._mxst
# Mean Frequencies :
x, _ = binArray(x, self._fSplitIndex, axis=0)
# Get the baseline and set to same shape of x:
xFm = np.mean(x[:, bsl[0]:bsl[1], :], 1)
# Mean time :
if self._window is not None:
x, _ = binArray(x, self._window, axis=1)
# Repeat baseline:
baseline = np.tile(xFm[:, np.newaxis, :], [1, x.shape[1], 1])
# Get shape of x:
nf, npts, nt = x.shape
pvalues = np.ones((nf, npts))
# Switch between methods:
# -> Permutations
# Loops on time and matrix for frequency (avoid RAM usage but increase speed)
if statmeth == 'permutation':
# Get metric:
fcn = perm_metric(self._metric)
# Apply metric to x and baseline:
xN = fcn(x, baseline).mean(axis=2)
# For each time points:
for pts in range(npts):
# Randomly swap x // baseline :
perm = perm_swap(x[:, pts, :],
baseline[:, pts, :],
n_perm=n_perm,
axis=1,
rndstate=0 + pts)[0]
# Normalize permutations by baline:
perm = fcn(perm, baseline[:, pts, :]).mean(2)
# Maximum stat (correct through frequencies):
if maxst:
perm = maxstat(perm, axis=1)
# Get pvalues :
pvalues[:, pts] = perm_2pvalue(xN[:, pts], perm, n_perm, tail=tail)
# -> Wilcoxon // Kruskal-Wallis:
else:
# Get the method:
if statmeth == 'wilcoxon':
def fcn(a, b):
return wilcoxon(a, b)[1]
elif statmeth == 'kruskal':
def fcn(a, b):
return kruskal(a, b)[1]
# Apply:
ite = product(range(nf), range(npts))
for k, i in ite:
pvalues[k, i] = fcn(x[k, i, :], xFm[k, :])
return pvalues
def get(self, xpha, xamp, n_perm=200, p=0.05, matricial=False, n_jobs=-1):
"""Get the normalized cfc mesure between an xpha and xamp signals.
Args:
xpha: array
Signal for phase. The shape of xpha should be :
(n_electrodes x n_pts x n_trials)
xamp: array
Signal for amplitude. The shape of xamp should be :
(n_electrodes x n_pts x n_trials)
Kargs:
n_perm: integer, optional, [def: 200]
Number of permutations for normalizing the cfc mesure.
p: float, optional, [def: 0.05]
p-value for the statistical method of Ozkurt 2012.
matricial: bool, optional, [def: False]
Some methods can work in matricial computation. This can lead
to a 10x or 30x time faster. But, please, monitor your RAM usage
beacause this parameter can use a lot of RAM. So, turn this parameter
in case of small computation.
n_jobs: integer, optional, [def: -1]
Control the number of jobs for parallel computing. Use 1, 2, ..
depending of your number or cores. -1 for all the cores.
If the same signal is used (example : xpha=x and xamp=x), this mean
the program compute a local cfc.
Returns:
ncfc: array
The cfc mesure of size :
(n_amplitude x n_phase x n_electrodes x n_windows x n_trials)
pvalue: array
The associated p-values of size :
(n_amplitude x n_phase x n_electrodes x n_windows)
"""
# Check the inputs variables :
xpha, xamp = _cfcCheck(xpha, xamp, self._npts)
self.n_perm = n_perm
self._matricial = matricial
if n_perm != 0:
self.p = 1/n_perm
else:
self.p = None
N = xpha.shape[0]
# Manage jobs repartition :
if (N < cpu_count()) and (n_jobs != 1):
surJob = n_jobs
elecJob = 1
elif (N >= cpu_count()) and (n_jobs != 1):
surJob = 1
elecJob = n_jobs
else:
surJob, elecJob = 1, 1
# Get the unormalized cfc and surogates:
cfcsu = Parallel(n_jobs=elecJob)(delayed(_cfcFiltSuro)(
xpha[k, ...], xamp[k, ...], surJob, self) for k in range(N))
uCfc, Suro, mSuro, stdSuro = zip(*cfcsu)
uCfc = np.array(uCfc)
# Permutations ans stat:
if (self.Id[0] is not '5'):
# Compute permutations :
if (self.n_perm is not 0) and (self.Id[1] is not '0'):
Suro, mSuro, stdSuro = np.array(
Suro), np.array(mSuro), np.array(stdSuro)
# Normalize each cfc:
_, _, Norm, _, _, _ = CfcSettings(self.Id)
nCfc = Norm(uCfc, mSuro, stdSuro)
# Confidence interval :
pvalue = perm_2pvalue(uCfc.mean(2), np.rollaxis(Suro.mean(2), 4),
self.n_perm, tail=1)
return nCfc.transpose(3, 4, 0, 1, 2), pvalue.transpose(2, 3, 0, 1)
else:
return uCfc.transpose(3, 4, 0, 1, 2), None
elif self.Id[0] is '5':
# Ozkurt threshold :
xlim = (erfinv(1-p)**2)
# Set to zero non-significant values:
idxUn = np.where(uCfc <= 2*xlim)
uCfc[idxUn] = 0
return uCfc.transpose(3, 4, 0, 1, 2), None
def fit(self, x, mf=False, center=False, grp=None, method="bino", n_perm=200, rndstate=0, n_jobs=-1):
"""Apply the classification and cross-validation objects to the array x.
Args:
x: list
List of dataset for each subject. All the dataset in the list
should have the same number of columns but the number of lines
could be diffrent for each subject and must correspond to the
same number of lines each each label vector of y.
Kargs:
mf: bool, optional, [def: False]
If mf=False, the returned decoding accuracy (da) will have a
shape of (1, rep) where rep, is the number of repetitions.
This mean that all the features are used together. If mf=True,
da.shape = (M, rep), where M is the number of columns of x.
center: optional, bool, [def: False]
Normalize fatures with a zero mean by substracting then dividing
by the mean. The center parameter should be set to True if the
classifier is a svm.
grp: array, optional, [def: None]
If mf=True, the grp parameter allow to define group of features.
If x.shape = (N, 5) and grp=np.array([0,0,1,2,1]), this mean that
3 groups of features will be considered : (0,1,2)
method: string, optional, [def: 'bino']
Four methods are implemented to test the statistical significiance
of the decoding accuracy :
- 'bino': binomial test
- 'label_rnd': randomly shuffle the labels
Methods 2 and 3 are based on permutations. They should provide
similar results. But 4 should be more conservative.
n_perm: integer, optional, [def: 200]
Number of permutations for the methods 2, 3 and 4
rndstate: integer, optional, [def: 0]
Fix the random state of the machine. Usefull to reproduce results.
n_jobs: integer, optional, [def: -1]
Control the number of jobs to cumpute the decoding accuracy. If
n_jobs = -1, all the jobs are used.
Return:
da: array
The decoding accuracy of shape n_repetitions x n_features
pvalue: array
Array of associated pvalue of shape n_features
daPerm: array
Array of all the decodings obtained for each permutations of shape
n_perm x n_features
.. rubric:: Footnotes
.. [#f8] `Ojala and Garriga, 2010 <http://www.jmlr.org/papers/volume11/ojala10a/ojala10a.pdf>`_
.. [#f9] `Combrisson and Jerbi, 2015 <http://www.ncbi.nlm.nih.gov/pubmed/25596422/>`_
"""
# Check x, y:
xbk = x.copy()
x, y, train, test = _checkXY(x, self._y, mf, grp, center, self)
nsuj, nfeat = x.shape
# Run classification:
da, ytrue, ypred = _fit(x, y, train, test, self, n_jobs)
# Get statistics:
# -------------------------------------------------------------
# Binomial :
# -------------------------------------------------------------
if method == "bino":
pvalue = bino_da2p(self._ry, da)
daperm = None
pperm = None
# -------------------------------------------------------------
# Permutations :
# -------------------------------------------------------------
# -> Shuffle the labels :
elif method == "label_rnd":
y_sh = [
_checkXY(xbk, [np.random.permutation(i) for i in self._y], mf, grp, center, self)[1]
for k in range(n_perm)
]
cvs = Parallel(n_jobs=n_jobs)(delayed(_fit)(x, y_sh[k], train, test, self, 1) for k in range(n_perm))
# Reconstruct daperm and get the associated p-value:
daperm, _, _ = zip(*cvs)
daperm = np.array(daperm).reshape(n_perm, nfeat)
pvalue = perm_2pvalue(da, daperm, n_perm, tail=1)
pperm = pvalue
else:
raise ValueError("No statistical method " + method + " found")
# Try to get featinfo:
try:
if grp is not None:
grp = uorderlst(grp)
else:
grp = np.arange(nfeat)
self.info.featinfo = self.info._featinfo(self._clf, self._cv, da[:, np.newaxis], grp=grp, pperm=pperm)
except:
pass
return da, pvalue, daperm
def _evalstat(self, x, bsl):
"""Statistical evaluation of features
[x] = [xn] = (nFce, npts, nTrials)
"""
# Unpack variables:
statmeth = self._statmeth
n_perm = self._n_perm
tail = self._2t
maxst = self._mxst
# Mean Frequencies :
x, _ = binArray(x, self._fSplitIndex, axis=0)
# Get the baseline and set to same shape of x:
xFm = np.mean(x[:, bsl[0]:bsl[1], :], 1)
# Mean time :
if self._window is not None:
x, _ = binArray(x, self._window, axis=1)
# Repeat baseline:
baseline = np.tile(xFm[:, np.newaxis, :], [1, x.shape[1], 1])
# Get shape of x:
nf, npts, nt = x.shape
pvalues = np.ones((nf, npts))
# Switch between methods:
# -> Permutations
# Loops on time and matrix for frequency (avoid RAM usage but increase speed)
if statmeth == 'permutation':
# Get metric:
fcn = perm_metric(self._metric)
# Apply metric to x and baseline:
xN = fcn(x, baseline).mean(axis=2)
# For each time points:
for pts in range(npts):
# Randomly swap x // baseline :
perm = perm_swap(x[:, pts, :], baseline[:, pts, :],
n_perm=n_perm, axis=1, rndstate=0+pts)[0]
# Normalize permutations by baline:
perm = fcn(perm, baseline[:, pts, :]).mean(2)
# Maximum stat (correct through frequencies):
if maxst:
perm = maxstat(perm, axis=1)
# Get pvalues :
pvalues[:, pts] = perm_2pvalue(xN[:, pts], perm, n_perm, tail=tail)
# -> Wilcoxon // Kruskal-Wallis:
else:
# Get the method:
if statmeth == 'wilcoxon':
def fcn(a, b): return wilcoxon(a, b)[1]
elif statmeth == 'kruskal':
def fcn(a, b): return kruskal(a, b)[1]
# Apply:
ite = product(range(nf), range(npts))
for k, i in ite:
pvalues[k, i] = fcn(x[k, i, :], xFm[k, :])
return pvalues
