_id
stringlengths 2
7
| title
stringlengths 1
88
| partition
stringclasses 3
values | text
stringlengths 75
19.8k
| language
stringclasses 1
value | meta_information
dict |
|---|---|---|---|---|---|
q500
|
load_parcellation_coords
|
train
|
def load_parcellation_coords(parcellation_name):
"""
Loads coordinates of included parcellations.
Parameters
----------
parcellation_name : str
options: 'gordon2014_333', 'power2012_264', 'shen2013_278'.
Returns
-------
parc : array
parcellation cordinates
"""
path = tenetopath[0] + '/data/parcellation/' + parcellation_name + '.csv'
parc = np.loadtxt(path, skiprows=1, delimiter=',', usecols=[1, 2, 3])
return parc
|
python
|
{
"resource": ""
}
|
q501
|
create_traj_ranges
|
train
|
def create_traj_ranges(start, stop, N):
"""
Fills in the trajectory range.
# Adapted from https://stackoverflow.com/a/40624614
"""
steps = (1.0/(N-1)) * (stop - start)
if np.isscalar(steps):
return steps*np.arange(N) + start
else:
return steps[:, None]*np.arange(N) + start[:, None]
|
python
|
{
"resource": ""
}
|
q502
|
get_dimord
|
train
|
def get_dimord(measure, calc=None, community=None):
"""
Get the dimension order of a network measure.
Parameters
----------
measure : str
Name of funciton in teneto.networkmeasures.
calc : str, default=None
Calc parameter for the function
community : bool, default=None
If not null, then community property is assumed to be believed.
Returns
-------
dimord : str
Dimension order. So "node,node,time" would define the dimensions of the network measure.
"""
if not calc:
calc = ''
else:
calc = '_' + calc
if not community:
community = ''
else:
community = 'community'
if 'community' in calc and 'community' in community:
community = ''
if calc == 'community_avg' or calc == 'community_pairs':
community = ''
dimord_dict = {
'temporal_closeness_centrality': 'node',
'temporal_degree_centrality': 'node',
'temporal_degree_centralit_avg': 'node',
'temporal_degree_centrality_time': 'node,time',
'temporal_efficiency': 'global',
'temporal_efficiency_global': 'global',
'temporal_efficiency_node': 'node',
'temporal_efficiency_to': 'node',
'sid_global': 'global,time',
'community_pairs': 'community,community,time',
'community_avg': 'community,time',
'sid': 'community,community,time',
'reachability_latency_global': 'global',
'reachability_latency': 'global',
'reachability_latency_node': 'node',
'fluctuability': 'node',
'fluctuability_global': 'global',
'bursty_coeff': 'edge,edge',
'bursty_coeff_edge': 'edge,edge',
'bursty_coeff_node': 'node',
'bursty_coeff_meanEdgePerNode': 'node',
'volatility_global': 'time',
}
if measure + calc + community in dimord_dict:
return dimord_dict[measure + calc + community]
else:
print('WARNINGL: get_dimord() returned unknown dimension labels')
return 'unknown'
|
python
|
{
"resource": ""
}
|
q503
|
create_supraadjacency_matrix
|
train
|
def create_supraadjacency_matrix(tnet, intersliceweight=1):
"""
Returns a supraadjacency matrix from a temporal network structure
Parameters
--------
tnet : TemporalNetwork
Temporal network (any network type)
intersliceweight : int
Weight that links the same node from adjacent time-points
Returns
--------
supranet : dataframe
Supraadjacency matrix
"""
newnetwork = tnet.network.copy()
newnetwork['i'] = (tnet.network['i']) + \
((tnet.netshape[0]) * (tnet.network['t']))
newnetwork['j'] = (tnet.network['j']) + \
((tnet.netshape[0]) * (tnet.network['t']))
if 'weight' not in newnetwork.columns:
newnetwork['weight'] = 1
newnetwork.drop('t', axis=1, inplace=True)
timepointconns = pd.DataFrame()
timepointconns['i'] = np.arange(0, (tnet.N*tnet.T)-tnet.N)
timepointconns['j'] = np.arange(tnet.N, (tnet.N*tnet.T))
timepointconns['weight'] = intersliceweight
supranet = pd.concat([newnetwork, timepointconns]).reset_index(drop=True)
return supranet
|
python
|
{
"resource": ""
}
|
q504
|
tnet_to_nx
|
train
|
def tnet_to_nx(df, t=None):
"""
Creates undirected networkx object
"""
if t is not None:
df = get_network_when(df, t=t)
if 'weight' in df.columns:
nxobj = nx.from_pandas_edgelist(
df, source='i', target='j', edge_attr='weight')
else:
nxobj = nx.from_pandas_edgelist(df, source='i', target='j')
return nxobj
|
python
|
{
"resource": ""
}
|
q505
|
temporal_louvain
|
train
|
def temporal_louvain(tnet, resolution=1, intersliceweight=1, n_iter=100, negativeedge='ignore', randomseed=None, consensus_threshold=0.5, temporal_consensus=True, njobs=1):
r"""
Louvain clustering for a temporal network.
Parameters
-----------
tnet : array, dict, TemporalNetwork
Input network
resolution : int
resolution of Louvain clustering ($\gamma$)
intersliceweight : int
interslice weight of multilayer clustering ($\omega$). Must be positive.
n_iter : int
Number of iterations to run louvain for
randomseed : int
Set for reproduceability
negativeedge : str
If there are negative edges, what should be done with them.
Options: 'ignore' (i.e. set to 0). More options to be added.
consensus : float (0.5 default)
When creating consensus matrix to average over number of iterations, keep values when the consensus is this amount.
Returns
-------
communities : array (node,time)
node,time array of community assignment
Notes
-------
References
----------
"""
tnet = process_input(tnet, ['C', 'G', 'TN'], 'TN')
# Divide resolution by the number of timepoints
resolution = resolution / tnet.T
supranet = create_supraadjacency_matrix(
tnet, intersliceweight=intersliceweight)
if negativeedge == 'ignore':
supranet = supranet[supranet['weight'] > 0]
nxsupra = tnet_to_nx(supranet)
np.random.seed(randomseed)
while True:
comtmp = []
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(_run_louvain, nxsupra, resolution, tnet.N, tnet.T) for n in range(n_iter)}
for j in as_completed(job):
comtmp.append(j.result())
comtmp = np.stack(comtmp)
comtmp = comtmp.transpose()
comtmp = np.reshape(comtmp, [tnet.N, tnet.T, n_iter], order='F')
if n_iter == 1:
break
nxsupra_old = nxsupra
nxsupra = make_consensus_matrix(comtmp, consensus_threshold)
# If there was no consensus, there are no communities possible, return
if nxsupra is None:
break
if (nx.to_numpy_array(nxsupra, nodelist=np.arange(tnet.N*tnet.T)) == nx.to_numpy_array(nxsupra_old, nodelist=np.arange(tnet.N*tnet.T))).all():
break
communities = comtmp[:, :, 0]
if temporal_consensus == True:
communities = make_temporal_consensus(communities)
return communities
|
python
|
{
"resource": ""
}
|
q506
|
make_temporal_consensus
|
train
|
def make_temporal_consensus(com_membership):
r"""
Matches community labels accross time-points
Jaccard matching is in a greedy fashiong. Matching the largest community at t with the community at t-1.
Parameters
----------
com_membership : array
Shape should be node, time.
Returns
-------
D : array
temporal consensus matrix using Jaccard distance
"""
com_membership = np.array(com_membership)
# make first indicies be between 0 and 1.
com_membership[:, 0] = clean_community_indexes(com_membership[:, 0])
# loop over all timepoints, get jacccard distance in greedy manner for largest community to time period before
for t in range(1, com_membership.shape[1]):
ct, counts_t = np.unique(com_membership[:, t], return_counts=True)
ct = ct[np.argsort(counts_t)[::-1]]
c1back = np.unique(com_membership[:, t-1])
new_index = np.zeros(com_membership.shape[0])
for n in ct:
if len(c1back) > 0:
d = np.ones(int(c1back.max())+1)
for m in c1back:
v1 = np.zeros(com_membership.shape[0])
v2 = np.zeros(com_membership.shape[0])
v1[com_membership[:, t] == n] = 1
v2[com_membership[:, t-1] == m] = 1
d[int(m)] = jaccard(v1, v2)
bestval = np.argmin(d)
else:
bestval = new_index.max() + 1
new_index[com_membership[:, t] == n] = bestval
c1back = np.array(np.delete(c1back, np.where(c1back == bestval)))
com_membership[:, t] = new_index
return com_membership
|
python
|
{
"resource": ""
}
|
q507
|
flexibility
|
train
|
def flexibility(communities):
"""
Amount a node changes community
Parameters
----------
communities : array
Community array of shape (node,time)
Returns
--------
flex : array
Size with the flexibility of each node.
Notes
-----
Flexbility calculates the number of times a node switches its community label during a time series. It is normalized by the number of possible changes which could occur. It is important to make sure that the different community labels accross time points are not artbirary.
References
-----------
Bassett, DS, Wymbs N, Porter MA, Mucha P, Carlson JM, Grafton ST. Dynamic reconfiguration of human brain networks during learning. PNAS, 2011, 108(18):7641-6.
"""
# Preallocate
flex = np.zeros(communities.shape[0])
# Go from the second time point to last, compare with time-point before
for t in range(1, communities.shape[1]):
flex[communities[:, t] != communities[:, t-1]] += 1
# Normalize
flex = flex / (communities.shape[1] - 1)
return flex
|
python
|
{
"resource": ""
}
|
q508
|
load_tabular_file
|
train
|
def load_tabular_file(fname, return_meta=False, header=True, index_col=True):
"""
Given a file name loads as a pandas data frame
Parameters
----------
fname : str
file name and path. Must be tsv.
return_meta :
header : bool (default True)
if there is a header in the tsv file, true will use first row in file.
index_col : bool (default None)
if there is an index column in the csv or tsv file, true will use first row in file.
Returns
-------
df : pandas
The loaded file
info : pandas, if return_meta=True
Meta infomration in json file (if specified)
"""
if index_col:
index_col = 0
else:
index_col = None
if header:
header = 0
else:
header = None
df = pd.read_csv(fname, header=header, index_col=index_col, sep='\t')
if return_meta:
json_fname = fname.replace('tsv', 'json')
meta = pd.read_json(json_fname)
return df, meta
else:
return df
|
python
|
{
"resource": ""
}
|
q509
|
get_sidecar
|
train
|
def get_sidecar(fname, allowedfileformats='default'):
"""
Loads sidecar or creates one
"""
if allowedfileformats == 'default':
allowedfileformats = ['.tsv', '.nii.gz']
for f in allowedfileformats:
fname = fname.split(f)[0]
fname += '.json'
if os.path.exists(fname):
with open(fname) as fs:
sidecar = json.load(fs)
else:
sidecar = {}
if 'filestatus' not in sidecar:
sidecar['filestatus'] = {}
sidecar['filestatus']['reject'] = False
sidecar['filestatus']['reason'] = []
return sidecar
|
python
|
{
"resource": ""
}
|
q510
|
process_exclusion_criteria
|
train
|
def process_exclusion_criteria(exclusion_criteria):
"""
Parses an exclusion critera string to get the function and threshold.
Parameters
----------
exclusion_criteria : list
list of strings where each string is of the format [relation][threshold]. E.g. \'<0.5\' or \'>=1\'
Returns
-------
relfun : list
list of numpy functions for the exclusion criteria
threshold : list
list of floats for threshold for each relfun
"""
relfun = []
threshold = []
for ec in exclusion_criteria:
if ec[0:2] == '>=':
relfun.append(np.greater_equal)
threshold.append(float(ec[2:]))
elif ec[0:2] == '<=':
relfun.append(np.less_equal)
threshold.append(float(ec[2:]))
elif ec[0] == '>':
relfun.append(np.greater)
threshold.append(float(ec[1:]))
elif ec[0] == '<':
relfun.append(np.less)
threshold.append(float(ec[1:]))
else:
raise ValueError('exclusion crieria must being with >,<,>= or <=')
return relfun, threshold
|
python
|
{
"resource": ""
}
|
q511
|
reachability_latency
|
train
|
def reachability_latency(tnet=None, paths=None, rratio=1, calc='global'):
"""
Reachability latency. This is the r-th longest temporal path.
Parameters
---------
data : array or dict
Can either be a network (graphlet or contact), binary unidrected only. Alternative can be a paths dictionary (output of teneto.networkmeasure.shortest_temporal_path)
rratio: float (default: 1)
reachability ratio that the latency is calculated in relation to.
Value must be over 0 and up to 1.
1 (default) - all nodes must be reached.
Other values (e.g. .5 imply that 50% of nodes are reached)
This is rounded to the nearest node inter.
E.g. if there are 6 nodes [1,2,3,4,5,6], it will be node 4 (due to round upwards)
calc : str
what to calculate. Alternatives: 'global' entire network; 'nodes': for each node.
Returns
--------
reach_lat : array
Reachability latency
Notes
------
Reachability latency calculates the time it takes for the paths.
"""
if tnet is not None and paths is not None:
raise ValueError('Only network or path input allowed.')
if tnet is None and paths is None:
raise ValueError('No input.')
# if shortest paths are not calculated, calculate them
if tnet is not None:
paths = shortest_temporal_path(tnet)
pathmat = np.zeros([paths[['from', 'to']].max().max(
)+1, paths[['from', 'to']].max().max()+1, paths[['t_start']].max().max()+1]) * np.nan
pathmat[paths['from'].values, paths['to'].values,
paths['t_start'].values] = paths['temporal-distance']
netshape = pathmat.shape
edges_to_reach = netshape[0] - np.round(netshape[0] * rratio)
reach_lat = np.zeros([netshape[1], netshape[2]]) * np.nan
for t_ind in range(0, netshape[2]):
paths_sort = -np.sort(-pathmat[:, :, t_ind], axis=1)
reach_lat[:, t_ind] = paths_sort[:, edges_to_reach]
if calc == 'global':
reach_lat = np.nansum(reach_lat)
reach_lat = reach_lat / ((netshape[0]) * netshape[2])
elif calc == 'nodes':
reach_lat = np.nansum(reach_lat, axis=1)
reach_lat = reach_lat / (netshape[2])
return reach_lat
|
python
|
{
"resource": ""
}
|
q512
|
recruitment
|
train
|
def recruitment(temporalcommunities, staticcommunities):
"""
Calculates recruitment coefficient for each node. Recruitment coefficient is the average probability of nodes from the
same static communities being in the same temporal communities at other time-points or during different tasks.
Parameters:
------------
temporalcommunities : array
temporal communities vector (node,time)
staticcommunities : array
Static communities vector for each node
Returns:
-------
Rcoeff : array
recruitment coefficient for each node
References:
-----------
Danielle S. Bassett, Muzhi Yang, Nicholas F. Wymbs, Scott T. Grafton.
Learning-Induced Autonomy of Sensorimotor Systems. Nat Neurosci. 2015 May;18(5):744-51.
Marcelo Mattar, Michael W. Cole, Sharon Thompson-Schill, Danielle S. Bassett. A Functional
Cartography of Cognitive Systems. PLoS Comput Biol. 2015 Dec 2;11(12):e1004533.
"""
# make sure the static and temporal communities have the same number of nodes
if staticcommunities.shape[0] != temporalcommunities.shape[0]:
raise ValueError(
'Temporal and static communities have different dimensions')
alleg = allegiance(temporalcommunities)
Rcoeff = np.zeros(len(staticcommunities))
for i, statcom in enumerate(staticcommunities):
Rcoeff[i] = np.mean(alleg[i, staticcommunities == statcom])
return Rcoeff
|
python
|
{
"resource": ""
}
|
q513
|
integration
|
train
|
def integration(temporalcommunities, staticcommunities):
"""
Calculates the integration coefficient for each node. Measures the average probability
that a node is in the same community as nodes from other systems.
Parameters:
------------
temporalcommunities : array
temporal communities vector (node,time)
staticcommunities : array
Static communities vector for each node
Returns:
-------
Icoeff : array
integration coefficient for each node
References:
----------
Danielle S. Bassett, Muzhi Yang, Nicholas F. Wymbs, Scott T. Grafton.
Learning-Induced Autonomy of Sensorimotor Systems. Nat Neurosci. 2015 May;18(5):744-51.
Marcelo Mattar, Michael W. Cole, Sharon Thompson-Schill, Danielle S. Bassett.
A Functional Cartography of Cognitive Systems. PLoS Comput Biol. 2015 Dec
2;11(12):e1004533.
"""
# make sure the static and temporal communities have the same number of nodes
if staticcommunities.shape[0] != temporalcommunities.shape[0]:
raise ValueError(
'Temporal and static communities have different dimensions')
alleg = allegiance(temporalcommunities)
Icoeff = np.zeros(len(staticcommunities))
# calc integration for each node
for i, statcom in enumerate(len(staticcommunities)):
Icoeff[i] = np.mean(alleg[i, staticcommunities != statcom])
return Icoeff
|
python
|
{
"resource": ""
}
|
q514
|
intercontacttimes
|
train
|
def intercontacttimes(tnet):
"""
Calculates the intercontacttimes of each edge in a network.
Parameters
-----------
tnet : array, dict
Temporal network (craphlet or contact). Nettype: 'bu', 'bd'
Returns
---------
contacts : dict
Intercontact times as numpy array in dictionary. contacts['intercontacttimes']
Notes
------
The inter-contact times is calculated by the time between consequecutive "active" edges (where active means
that the value is 1 in a binary network).
Examples
--------
This example goes through how inter-contact times are calculated.
>>> import teneto
>>> import numpy as np
Make a network with 2 nodes and 4 time-points with 4 edges spaced out.
>>> G = np.zeros([2,2,10])
>>> edge_on = [1,3,5,9]
>>> G[0,1,edge_on] = 1
The network visualised below make it clear what the inter-contact times are between the two nodes:
.. plot::
import teneto
import numpy as np
import matplotlib.pyplot as plt
G = np.zeros([2,2,10])
edge_on = [1,3,5,9]
G[0,1,edge_on] = 1
fig, ax = plt.subplots(1, figsize=(4,2))
teneto.plot.slice_plot(G, ax=ax, cmap='Pastel2')
ax.set_ylim(-0.25, 1.25)
plt.tight_layout()
fig.show()
Calculating the inter-contact times of these edges becomes: 2,2,4 between nodes 0 and 1.
>>> ict = teneto.networkmeasures.intercontacttimes(G)
The function returns a dictionary with the icts in the key: intercontacttimes. This is of the size NxN.
So the icts between nodes 0 and 1 are found by:
>>> ict['intercontacttimes'][0,1]
array([2, 2, 4])
"""
# Process input
tnet = process_input(tnet, ['C', 'G', 'TN'], 'TN')
if tnet.nettype[0] == 'w':
print('WARNING: assuming connections to be binary when computing intercontacttimes')
# Each time series is padded with a 0 at the start and end. Then t[0:-1]-[t:].
# Then discard the noninformative ones (done automatically)
# Finally return back as np array
contacts = np.array([[None] * tnet.netshape[0]] * tnet.netshape[0])
if tnet.nettype[1] == 'u':
for i in range(0, tnet.netshape[0]):
for j in range(i + 1, tnet.netshape[0]):
edge_on = tnet.get_network_when(i=i, j=j)['t'].values
if len(edge_on) > 0:
edge_on_diff = edge_on[1:] - edge_on[:-1]
contacts[i, j] = np.array(edge_on_diff)
contacts[j, i] = np.array(edge_on_diff)
else:
contacts[i, j] = []
contacts[j, i] = []
elif tnet.nettype[1] == 'd':
for i in range(0, tnet.netshape[0]):
for j in range(0, tnet.netshape[0]):
edge_on = tnet.get_network_when(i=i, j=j)['t'].values
if len(edge_on) > 0:
edge_on_diff = edge_on[1:] - edge_on[:-1]
contacts[i, j] = np.array(edge_on_diff)
else:
contacts[i, j] = []
out = {}
out['intercontacttimes'] = contacts
out['nettype'] = tnet.nettype
return out
|
python
|
{
"resource": ""
}
|
q515
|
gen_report
|
train
|
def gen_report(report, sdir='./', report_name='report.html'):
"""
Generates report of derivation and postprocess steps in teneto.derive
"""
# Create report directory
if not os.path.exists(sdir):
os.makedirs(sdir)
# Add a slash to file directory if not included to avoid DirNameFleName
# instead of DirName/FileName being creaated
if sdir[-1] != '/':
sdir += '/'
report_html = '<html><body>'
if 'method' in report.keys():
report_html += "<h1>Method: " + report['method'] + "</h1><p>"
for i in report[report['method']]:
if i == 'taper_window':
fig, ax = plt.subplots(1)
ax.plot(report[report['method']]['taper_window'],
report[report['method']]['taper'])
ax.set_xlabel('Window (time). 0 in middle of window.')
ax.set_title(
'Taper from ' + report[report['method']]['distribution'] + ' distribution (PDF).')
fig.savefig(sdir + 'taper.png')
report_html += "<img src='./taper.png' width=500>" + "<p>"
else:
report_html += "- <b>" + i + "</b>: " + \
str(report[report['method']][i]) + "<br>"
if 'postprocess' in report.keys():
report_html += "<p><h2>Postprocessing:</h2><p>"
report_html += "<b>Pipeline: </b>"
for i in report['postprocess']:
report_html += " " + i + ","
for i in report['postprocess']:
report_html += "<p><h3>" + i + "</h3><p>"
for j in report[i]:
if j == 'lambda':
report_html += "- <b>" + j + "</b>: " + "<br>"
lambda_val = np.array(report['boxcox']['lambda'])
fig, ax = plt.subplots(1)
ax.hist(lambda_val[:, -1])
ax.set_xlabel('lambda')
ax.set_ylabel('frequency')
ax.set_title('Histogram of lambda parameter')
fig.savefig(sdir + 'boxcox_lambda.png')
report_html += "<img src='./boxcox_lambda.png' width=500>" + "<p>"
report_html += "Data located in " + sdir + "boxcox_lambda.csv <p>"
np.savetxt(sdir + "boxcox_lambda.csv",
lambda_val, delimiter=",")
else:
report_html += "- <b>" + j + "</b>: " + \
str(report[i][j]) + "<br>"
report_html += '</body></html>'
with open(sdir + report_name, 'w') as file:
file.write(report_html)
file.close()
|
python
|
{
"resource": ""
}
|
q516
|
TenetoBIDS.add_history
|
train
|
def add_history(self, fname, fargs, init=0):
"""
Adds a processing step to TenetoBIDS.history.
"""
if init == 1:
self.history = []
self.history.append([fname, fargs])
|
python
|
{
"resource": ""
}
|
q517
|
TenetoBIDS.derive_temporalnetwork
|
train
|
def derive_temporalnetwork(self, params, update_pipeline=True, tag=None, njobs=1, confound_corr_report=True):
"""
Derive time-varying connectivity on the selected files.
Parameters
----------
params : dict.
See teneto.timeseries.derive_temporalnetwork for the structure of the param dictionary. Assumes dimord is time,node (output of other TenetoBIDS funcitons)
update_pipeline : bool
If true, the object updates the selected files with those derived here.
njobs : int
How many parallel jobs to run
confound_corr_report : bool
If true, histograms and summary statistics of TVC and confounds are plotted in a report directory.
tag : str
any additional tag that will be placed in the saved file name. Will be placed as 'desc-[tag]'
Returns
-------
dfc : files
saved in .../derivatives/teneto/sub-xxx/tvc/..._tvc.npy
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
files = self.get_selected_files(quiet=1)
confound_files = self.get_selected_files(quiet=1, pipeline='confound')
if confound_files:
confounds_exist = True
else:
confounds_exist = False
if not confound_corr_report:
confounds_exist = False
if not tag:
tag = ''
else:
tag = 'desc-' + tag
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(self._derive_temporalnetwork, f, i, tag, params,
confounds_exist, confound_files) for i, f in enumerate(files) if f}
for j in as_completed(job):
j.result()
if update_pipeline == True:
if not self.confound_pipeline and len(self.get_selected_files(quiet=1, pipeline='confound')) > 0:
self.set_confound_pipeline = self.pipeline
self.set_pipeline('teneto_' + teneto.__version__)
self.set_pipeline_subdir('tvc')
self.set_bids_suffix('tvcconn')
|
python
|
{
"resource": ""
}
|
q518
|
TenetoBIDS.make_functional_connectivity
|
train
|
def make_functional_connectivity(self, njobs=None, returngroup=False, file_hdr=None, file_idx=None):
"""
Makes connectivity matrix for each of the subjects.
Parameters
----------
returngroup : bool, default=False
If true, returns the group average connectivity matrix.
njobs : int
How many parallel jobs to run
file_idx : bool
Default False, true if to ignore index column in loaded file.
file_hdr : bool
Default False, true if to ignore header row in loaded file.
Returns
-------
Saves data in derivatives/teneto_<version>/.../fc/
R_group : array
if returngroup is true, the average connectivity matrix is returned.
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
files = self.get_selected_files(quiet=1)
R_group = []
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(
self._run_make_functional_connectivity, f, file_hdr, file_idx) for f in files}
for j in as_completed(job):
R_group.append(j.result())
if returngroup:
# Fisher tranform -> mean -> inverse fisher tranform
R_group = np.tanh(np.mean(np.arctanh(np.array(R_group)), axis=0))
return np.array(R_group)
|
python
|
{
"resource": ""
}
|
q519
|
TenetoBIDS._save_namepaths_bids_derivatives
|
train
|
def _save_namepaths_bids_derivatives(self, f, tag, save_directory, suffix=None):
"""
Creates output directory and output name
Paramters
---------
f : str
input files, includes the file bids_suffix
tag : str
what should be added to f in the output file.
save_directory : str
additional directory that the output file should go in
suffix : str
add new suffix to data
Returns
-------
save_name : str
previous filename with new tag
save_dir : str
directory where it will be saved
base_dir : str
subjective base directory (i.e. derivatives/teneto/func[/anythingelse/])
"""
file_name = f.split('/')[-1].split('.')[0]
if tag != '':
tag = '_' + tag
if suffix:
file_name, _ = drop_bids_suffix(file_name)
save_name = file_name + tag
save_name += '_' + suffix
else:
save_name = file_name + tag
paths_post_pipeline = f.split(self.pipeline)
if self.pipeline_subdir:
paths_post_pipeline = paths_post_pipeline[1].split(self.pipeline_subdir)[
0]
else:
paths_post_pipeline = paths_post_pipeline[1].split(file_name)[0]
base_dir = self.BIDS_dir + '/derivatives/' + 'teneto_' + \
teneto.__version__ + '/' + paths_post_pipeline + '/'
save_dir = base_dir + '/' + save_directory + '/'
if not os.path.exists(save_dir):
# A case has happened where this has been done in parallel and an error was raised. So do try/except
try:
os.makedirs(save_dir)
except:
# Wait 2 seconds so that the error does not try and save something in the directory before it is created
time.sleep(2)
if not os.path.exists(self.BIDS_dir + '/derivatives/' + 'teneto_' + teneto.__version__ + '/dataset_description.json'):
try:
with open(self.BIDS_dir + '/derivatives/' + 'teneto_' + teneto.__version__ + '/dataset_description.json', 'w') as fs:
json.dump(self.tenetoinfo, fs)
except:
# Same as above, just in case parallel does duplicaiton
time.sleep(2)
return save_name, save_dir, base_dir
|
python
|
{
"resource": ""
}
|
q520
|
TenetoBIDS.get_tags
|
train
|
def get_tags(self, tag, quiet=1):
"""
Returns which tag alternatives can be identified in the BIDS derivatives structure.
"""
if not self.pipeline:
print('Please set pipeline first.')
self.get_pipeline_alternatives(quiet)
else:
if tag == 'sub':
datapath = self.BIDS_dir + '/derivatives/' + self.pipeline + '/'
tag_alternatives = [
f.split('sub-')[1] for f in os.listdir(datapath) if os.path.isdir(datapath + f) and 'sub-' in f]
elif tag == 'ses':
tag_alternatives = []
for sub in self.bids_tags['sub']:
tag_alternatives += [f.split('ses-')[1] for f in os.listdir(
self.BIDS_dir + '/derivatives/' + self.pipeline + '/' + 'sub-' + sub) if 'ses' in f]
tag_alternatives = set(tag_alternatives)
else:
files = self.get_selected_files(quiet=1)
tag_alternatives = []
for f in files:
f = f.split('.')[0]
f = f.split('/')[-1]
tag_alternatives += [t.split('-')[1]
for t in f.split('_') if t.split('-')[0] == tag]
tag_alternatives = set(tag_alternatives)
if quiet == 0:
print(tag + ' alternatives: ' + ', '.join(tag_alternatives))
return list(tag_alternatives)
|
python
|
{
"resource": ""
}
|
q521
|
TenetoBIDS.set_exclusion_file
|
train
|
def set_exclusion_file(self, confound, exclusion_criteria, confound_stat='mean'):
"""
Excludes subjects given a certain exclusion criteria.
Parameters
----------
confound : str or list
string or list of confound name(s) from confound files
exclusion_criteria : str or list
for each confound, an exclusion_criteria should be expressed as a string. It starts with >,<,>= or <= then the numerical threshold. Ex. '>0.2' will entail every subject with the avg greater than 0.2 of confound will be rejected.
confound_stat : str or list
Can be median, mean, std. How the confound data is aggregated (so if there is a meaasure per time-point, this is averaged over all time points. If multiple confounds specified, this has to be a list.).
Returns
--------
calls TenetoBIDS.set_bad_files with the files meeting the exclusion criteria.
"""
self.add_history(inspect.stack()[0][3], locals(), 1)
if isinstance(confound, str):
confound = [confound]
if isinstance(exclusion_criteria, str):
exclusion_criteria = [exclusion_criteria]
if isinstance(confound_stat, str):
confound_stat = [confound_stat]
if len(exclusion_criteria) != len(confound):
raise ValueError(
'Same number of confound names and exclusion criteria must be given')
if len(confound_stat) != len(confound):
raise ValueError(
'Same number of confound names and confound stats must be given')
relex, crit = process_exclusion_criteria(exclusion_criteria)
files = sorted(self.get_selected_files(quiet=1))
confound_files = sorted(
self.get_selected_files(quiet=1, pipeline='confound'))
files, confound_files = confound_matching(files, confound_files)
bad_files = []
bs = 0
foundconfound = []
foundreason = []
for s, cfile in enumerate(confound_files):
df = load_tabular_file(cfile, index_col=None)
found_bad_subject = False
for i, _ in enumerate(confound):
if confound_stat[i] == 'median':
if relex[i](df[confound[i]].median(), crit[i]):
found_bad_subject = True
elif confound_stat[i] == 'mean':
if relex[i](df[confound[i]].mean(), crit[i]):
found_bad_subject = True
elif confound_stat[i] == 'std':
if relex[i](df[i][confound[i]].std(), crit[i]):
found_bad_subject = True
if found_bad_subject:
foundconfound.append(confound[i])
foundreason.append(exclusion_criteria[i])
if found_bad_subject:
bad_files.append(files[s])
bs += 1
self.set_bad_files(
bad_files, reason='excluded file (confound over specfied stat threshold)')
for i, f in enumerate(bad_files):
sidecar = get_sidecar(f)
sidecar['file_exclusion'] = {}
sidecar['confound'] = foundconfound[i]
sidecar['threshold'] = foundreason[i]
for af in ['.tsv', '.nii.gz']:
f = f.split(af)[0]
f += '.json'
with open(f, 'w') as fs:
json.dump(sidecar, fs)
print('Removed ' + str(bs) + ' files from inclusion.')
|
python
|
{
"resource": ""
}
|
q522
|
TenetoBIDS.make_parcellation
|
train
|
def make_parcellation(self, parcellation, parc_type=None, parc_params=None, network='defaults', update_pipeline=True, removeconfounds=False, tag=None, njobs=None, clean_params=None, yeonetworkn=None):
"""
Reduces the data from voxel to parcellation space. Files get saved in a teneto folder in the derivatives with a roi tag at the end.
Parameters
-----------
parcellation : str
specify which parcellation that you would like to use. For MNI: 'power2012_264', 'gordon2014_333'. TAL: 'shen2013_278'
parc_type : str
can be 'sphere' or 'region'. If nothing is specified, the default for that parcellation will be used.
parc_params : dict
**kwargs for nilearn functions
network : str
if "defaults", it selects static parcellation, _if available_ (other options will be made available soon).
removeconfounds : bool
if true, regresses out confounds that are specfied in self.set_confounds with linear regression.
update_pipeline : bool
TenetoBIDS gets updated with the parcellated files being selected.
tag : str or list
any additional tag that must be in file name. After the tag there must either be a underscore or period (following bids).
clean_params : dict
**kwargs for nilearn function nilearn.signal.clean
yeonetworkn : int (7 or 17)
Only relevant for when parcellation is schaeffer2018. Use 7 or 17 template networks
njobs : n
number of processes to run. Overrides TenetoBIDS.njobs
Returns
-------
Files are saved in ./BIDS_dir/derivatives/teneto_<version>/.../parcellation/.
To load these files call TenetoBIDS.load_parcellation.
NOTE
----
These functions make use of nilearn. Please cite nilearn if used in a publicaiton.
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
parc_name = parcellation.split('_')[0].lower()
# Check confounds have been specified
if not self.confounds and removeconfounds:
raise ValueError(
'Specified confounds are not found. Make sure that you have run self.set_confunds([\'Confound1\',\'Confound2\']) first.')
# Check confounds have been specified
if update_pipeline == False and removeconfounds:
raise ValueError(
'Pipeline must be updated in order to remove confounds within this funciton.')
# In theory these should be the same. So at the moment, it goes through each element and checks they are matched.
# A matching algorithem may be needed if cases arise where this isnt the case
files = self.get_selected_files(quiet=1)
# Load network communities, if possible.
self.set_network_communities(parcellation, netn=yeonetworkn)
if not tag:
tag = ''
else:
tag = 'desc-' + tag
if not parc_params:
parc_params = {}
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(self._run_make_parcellation, f, i, tag, parcellation,
parc_name, parc_type, parc_params) for i, f in enumerate(files)}
for j in as_completed(job):
j.result()
if update_pipeline == True:
if not self.confound_pipeline and len(self.get_selected_files(quiet=1, pipeline='confound')) > 0:
self.set_confound_pipeline(self.pipeline)
self.set_pipeline('teneto_' + teneto.__version__)
self.set_pipeline_subdir('parcellation')
if tag:
self.set_bids_tags({'desc': tag.split('-')[1]})
self.set_bids_suffix('roi')
if removeconfounds:
self.removeconfounds(
clean_params=clean_params, transpose=None, njobs=njobs)
|
python
|
{
"resource": ""
}
|
q523
|
TenetoBIDS.communitydetection
|
train
|
def communitydetection(self, community_detection_params, community_type='temporal', tag=None, file_hdr=False, file_idx=False, njobs=None):
"""
Calls temporal_louvain_with_consensus on connectivity data
Parameters
----------
community_detection_params : dict
kwargs for detection. See teneto.communitydetection.louvain.temporal_louvain_with_consensus
community_type : str
Either 'temporal' or 'static'. If temporal, community is made per time-point for each timepoint.
file_idx : bool (default false)
if true, index column present in data and this will be ignored
file_hdr : bool (default false)
if true, header row present in data and this will be ignored
njobs : int
number of processes to run. Overrides TenetoBIDS.njobs
Note
----
All non-positive edges are made to zero.
Returns
-------
List of communities for each subject. Saved in BIDS_dir/derivatives/teneto/communitydetection/
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
if not tag:
tag = ''
else:
tag = 'desc-' + tag
if community_type == 'temporal':
files = self.get_selected_files(quiet=True)
# Run check to make sure files are tvc input
for f in files:
if 'tvc' not in f:
raise ValueError(
'tvc tag not found in filename. TVC data must be used in communitydetection (perhaps run TenetoBIDS.derive first?).')
elif community_type == 'static':
files = self.get_selected_files(
quiet=True, pipeline='functionalconnectivity')
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(self._run_communitydetection, f, community_detection_params, community_type, file_hdr,
file_idx, tag) for i, f in enumerate(files) if all([t + '_' in f or t + '.' in f for t in tag])}
for j in as_completed(job):
j.result()
|
python
|
{
"resource": ""
}
|
q524
|
TenetoBIDS.removeconfounds
|
train
|
def removeconfounds(self, confounds=None, clean_params=None, transpose=None, njobs=None, update_pipeline=True, overwrite=True, tag=None):
"""
Removes specified confounds using nilearn.signal.clean
Parameters
----------
confounds : list
List of confounds. Can be prespecified in set_confounds
clean_params : dict
Dictionary of kawgs to pass to nilearn.signal.clean
transpose : bool (default False)
Default removeconfounds works on time,node dimensions. Pass transpose=True to transpose pre and post confound removal.
njobs : int
Number of jobs. Otherwise tenetoBIDS.njobs is run.
update_pipeline : bool
update pipeline with '_clean' tag for new files created
overwrite : bool
tag : str
Returns
-------
Says all TenetBIDS.get_selected_files with confounds removed with _rmconfounds at the end.
Note
----
There may be some issues regarding loading non-cleaned data through the TenetoBIDS functions instead of the cleaned data. This depeneds on when you clean the data.
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
if not self.confounds and not confounds:
raise ValueError(
'Specified confounds are not found. Make sure that you have run self.set_confunds([\'Confound1\',\'Confound2\']) first or pass confounds as input to function.')
if not tag:
tag = ''
else:
tag = 'desc-' + tag
if confounds:
self.set_confounds(confounds)
files = sorted(self.get_selected_files(quiet=1))
confound_files = sorted(
self.get_selected_files(quiet=1, pipeline='confound'))
files, confound_files = confound_matching(files, confound_files)
if not clean_params:
clean_params = {}
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(
self._run_removeconfounds, f, confound_files[i], clean_params, transpose, overwrite, tag) for i, f in enumerate(files)}
for j in as_completed(job):
j.result()
self.set_pipeline('teneto_' + teneto.__version__)
self.set_bids_suffix('roi')
if tag:
self.set_bids_tags({'desc': tag.split('-')[1]})
|
python
|
{
"resource": ""
}
|
q525
|
TenetoBIDS.networkmeasures
|
train
|
def networkmeasures(self, measure=None, measure_params=None, tag=None, njobs=None):
"""
Calculates a network measure
For available funcitons see: teneto.networkmeasures
Parameters
----------
measure : str or list
Mame of function(s) from teneto.networkmeasures that will be run.
measure_params : dict or list of dctionaries)
Containing kwargs for the argument in measure.
See note regarding Communities key.
tag : str
Add additional tag to saved filenames.
Note
----
In measure_params, if communities can equal 'template', 'static', or 'temporal'.
These options must be precalculated. If template, Teneto tries to load default for parcellation. If static, loads static communities
in BIDS_dir/teneto_<version>/sub-.../func/communities/..._communitytype-static....npy. If temporal, loads static communities
in BIDS_dir/teneto_<version>/sub-.../func/communities/..._communitytype-temporal....npy
Returns
-------
Saves in ./BIDS_dir/derivatives/teneto/sub-NAME/func//temporalnetwork/MEASURE/
Load the measure with tenetoBIDS.load_network_measure
"""
if not njobs:
njobs = self.njobs
self.add_history(inspect.stack()[0][3], locals(), 1)
# measure can be string or list
if isinstance(measure, str):
measure = [measure]
# measure_params can be dictionaary or list of dictionaries
if isinstance(measure_params, dict):
measure_params = [measure_params]
if measure_params and len(measure) != len(measure_params):
raise ValueError('Number of identified measure_params (' + str(len(measure_params)) +
') differs from number of identified measures (' + str(len(measure)) + '). Leave black dictionary if default methods are wanted')
files = self.get_selected_files(quiet=1)
if not tag:
tag = ''
else:
tag = 'desc-' + tag
with ProcessPoolExecutor(max_workers=njobs) as executor:
job = {executor.submit(
self._run_networkmeasures, f, tag, measure, measure_params) for f in files}
for j in as_completed(job):
j.result()
|
python
|
{
"resource": ""
}
|
q526
|
TenetoBIDS.set_bids_suffix
|
train
|
def set_bids_suffix(self, bids_suffix):
"""
The last analysis step is the final tag that is present in files.
"""
self.add_history(inspect.stack()[0][3], locals(), 1)
self.bids_suffix = bids_suffix
|
python
|
{
"resource": ""
}
|
q527
|
TenetoBIDS.set_pipeline
|
train
|
def set_pipeline(self, pipeline):
"""
Specify the pipeline. See get_pipeline_alternatives to see what are avaialble. Input should be a string.
"""
self.add_history(inspect.stack()[0][3], locals(), 1)
if not os.path.exists(self.BIDS_dir + '/derivatives/' + pipeline):
print('Specified direvative directory not found.')
self.get_pipeline_alternatives()
else:
# Todo: perform check that pipeline is valid
self.pipeline = pipeline
|
python
|
{
"resource": ""
}
|
q528
|
TenetoBIDS.load_frompickle
|
train
|
def load_frompickle(cls, fname, reload_object=False):
"""
Loaded saved instance of
fname : str
path to pickle object (output of TenetoBIDS.save_aspickle)
reload_object : bool (default False)
reloads object by calling teneto.TenetoBIDS (some information lost, for development)
Returns
-------
self :
TenetoBIDS instance
"""
if fname[-4:] != '.pkl':
fname += '.pkl'
with open(fname, 'rb') as f:
tnet = pickle.load(f)
if reload_object:
reloadnet = teneto.TenetoBIDS(tnet.BIDS_dir, pipeline=tnet.pipeline, pipeline_subdir=tnet.pipeline_subdir, bids_tags=tnet.bids_tags, bids_suffix=tnet.bids_suffix,
bad_subjects=tnet.bad_subjects, confound_pipeline=tnet.confound_pipeline, raw_data_exists=tnet.raw_data_exists, njobs=tnet.njobs)
reloadnet.histroy = tnet.history
tnet = reloadnet
return tnet
|
python
|
{
"resource": ""
}
|
q529
|
temporal_closeness_centrality
|
train
|
def temporal_closeness_centrality(tnet=None, paths=None):
'''
Returns temporal closeness centrality per node.
Parameters
-----------
Input should be *either* tnet or paths.
data : array or dict
Temporal network input (graphlet or contact). nettype: 'bu', 'bd'.
paths : pandas dataframe
Output of TenetoBIDS.networkmeasure.shortest_temporal_paths
Returns
--------
:close: array
temporal closness centrality (nodal measure)
'''
if tnet is not None and paths is not None:
raise ValueError('Only network or path input allowed.')
if tnet is None and paths is None:
raise ValueError('No input.')
# if shortest paths are not calculated, calculate them
if tnet is not None:
paths = shortest_temporal_path(tnet)
pathmat = np.zeros([paths[['from', 'to']].max().max(
)+1, paths[['from', 'to']].max().max()+1, paths[['t_start']].max().max()+1]) * np.nan
pathmat[paths['from'].values, paths['to'].values,
paths['t_start'].values] = paths['temporal-distance']
closeness = np.nansum(1 / np.nanmean(pathmat, axis=2),
axis=1) / (pathmat.shape[1] - 1)
return closeness
|
python
|
{
"resource": ""
}
|
q530
|
flatten
|
train
|
def flatten(d, reducer='tuple', inverse=False):
"""Flatten dict-like object.
Parameters
----------
d: dict-like object
The dict that will be flattened.
reducer: {'tuple', 'path', function} (default: 'tuple')
The key joining method. If a function is given, the function will be
used to reduce.
'tuple': The resulting key will be tuple of the original keys
'path': Use ``os.path.join`` to join keys.
inverse: bool (default: False)
Whether you want invert the resulting key and value.
Returns
-------
flat_dict: dict
"""
if isinstance(reducer, str):
reducer = REDUCER_DICT[reducer]
flat_dict = {}
def _flatten(d, parent=None):
for key, value in six.viewitems(d):
flat_key = reducer(parent, key)
if isinstance(value, Mapping):
_flatten(value, flat_key)
else:
if inverse:
flat_key, value = value, flat_key
if flat_key in flat_dict:
raise ValueError("duplicated key '{}'".format(flat_key))
flat_dict[flat_key] = value
_flatten(d)
return flat_dict
|
python
|
{
"resource": ""
}
|
q531
|
nested_set_dict
|
train
|
def nested_set_dict(d, keys, value):
"""Set a value to a sequence of nested keys
Parameters
----------
d: Mapping
keys: Sequence[str]
value: Any
"""
assert keys
key = keys[0]
if len(keys) == 1:
if key in d:
raise ValueError("duplicated key '{}'".format(key))
d[key] = value
return
d = d.setdefault(key, {})
nested_set_dict(d, keys[1:], value)
|
python
|
{
"resource": ""
}
|
q532
|
unflatten
|
train
|
def unflatten(d, splitter='tuple', inverse=False):
"""Unflatten dict-like object.
Parameters
----------
d: dict-like object
The dict that will be unflattened.
splitter: {'tuple', 'path', function} (default: 'tuple')
The key splitting method. If a function is given, the function will be
used to split.
'tuple': Use each element in the tuple key as the key of the unflattened dict.
'path': Use ``pathlib.Path.parts`` to split keys.
inverse: bool (default: False)
Whether you want to invert the key and value before flattening.
Returns
-------
unflattened_dict: dict
"""
if isinstance(splitter, str):
splitter = SPLITTER_DICT[splitter]
unflattened_dict = {}
for flat_key, value in six.viewitems(d):
if inverse:
flat_key, value = value, flat_key
key_tuple = splitter(flat_key)
nested_set_dict(unflattened_dict, key_tuple, value)
return unflattened_dict
|
python
|
{
"resource": ""
}
|
q533
|
plot_track
|
train
|
def plot_track(track, filename=None, beat_resolution=None, downbeats=None,
preset='default', cmap='Blues', xtick='auto', ytick='octave',
xticklabel=True, yticklabel='auto', tick_loc=None,
tick_direction='in', label='both', grid='both',
grid_linestyle=':', grid_linewidth=.5):
"""
Plot the pianoroll or save a plot of the pianoroll.
Parameters
----------
filename :
The filename to which the plot is saved. If None, save nothing.
beat_resolution : int
The number of time steps used to represent a beat. Required and only
effective when `xtick` is 'beat'.
downbeats : list
An array that indicates whether the time step contains a downbeat
(i.e., the first time step of a bar).
preset : {'default', 'plain', 'frame'}
A string that indicates the preset theme to use.
- In 'default' preset, the ticks, grid and labels are on.
- In 'frame' preset, the ticks and grid are both off.
- In 'plain' preset, the x- and y-axis are both off.
cmap : `matplotlib.colors.Colormap`
The colormap to use in :func:`matplotlib.pyplot.imshow`. Defaults to
'Blues'. Only effective when `pianoroll` is 2D.
xtick : {'auto', 'beat', 'step', 'off'}
A string that indicates what to use as ticks along the x-axis. If
'auto' is given, automatically set to 'beat' if `beat_resolution` is
also given and set to 'step', otherwise. Defaults to 'auto'.
ytick : {'octave', 'pitch', 'off'}
A string that indicates what to use as ticks along the y-axis.
Defaults to 'octave'.
xticklabel : bool
Whether to add tick labels along the x-axis. Only effective when
`xtick` is not 'off'.
yticklabel : {'auto', 'name', 'number', 'off'}
If 'name', use octave name and pitch name (key name when `is_drum`
is True) as tick labels along the y-axis. If 'number', use pitch
number. If 'auto', set to 'name' when `ytick` is 'octave' and
'number' when `ytick` is 'pitch'. Defaults to 'auto'. Only effective
when `ytick` is not 'off'.
tick_loc : tuple or list
The locations to put the ticks. Availables elements are 'bottom',
'top', 'left' and 'right'. Defaults to ('bottom', 'left').
tick_direction : {'in', 'out', 'inout'}
A string that indicates where to put the ticks. Defaults to 'in'.
Only effective when one of `xtick` and `ytick` is on.
label : {'x', 'y', 'both', 'off'}
A string that indicates whether to add labels to the x-axis and
y-axis. Defaults to 'both'.
grid : {'x', 'y', 'both', 'off'}
A string that indicates whether to add grids to the x-axis, y-axis,
both or neither. Defaults to 'both'.
grid_linestyle : str
Will be passed to :meth:`matplotlib.axes.Axes.grid` as 'linestyle'
argument.
grid_linewidth : float
Will be passed to :meth:`matplotlib.axes.Axes.grid` as 'linewidth'
argument.
Returns
-------
fig : `matplotlib.figure.Figure` object
A :class:`matplotlib.figure.Figure` object.
ax : `matplotlib.axes.Axes` object
A :class:`matplotlib.axes.Axes` object.
"""
if not HAS_MATPLOTLIB:
raise ImportError("matplotlib package is required for plotting "
"supports.")
fig, ax = plt.subplots()
plot_pianoroll(ax, track.pianoroll, track.is_drum, beat_resolution,
downbeats, preset=preset, cmap=cmap, xtick=xtick,
ytick=ytick, xticklabel=xticklabel, yticklabel=yticklabel,
tick_loc=tick_loc, tick_direction=tick_direction,
label=label, grid=grid, grid_linestyle=grid_linestyle,
grid_linewidth=grid_linewidth)
if filename is not None:
plt.savefig(filename)
return fig, ax
|
python
|
{
"resource": ""
}
|
q534
|
Multitrack.append_track
|
train
|
def append_track(self, track=None, pianoroll=None, program=0, is_drum=False,
name='unknown'):
"""
Append a multitrack.Track instance to the track list or create a new
multitrack.Track object and append it to the track list.
Parameters
----------
track : pianoroll.Track
A :class:`pypianoroll.Track` instance to be appended to the track
list.
pianoroll : np.ndarray, shape=(n_time_steps, 128)
A pianoroll matrix. The first and second dimension represents time
and pitch, respectively. Available datatypes are bool, int and
float. Only effective when `track` is None.
program: int
A program number according to General MIDI specification [1].
Available values are 0 to 127. Defaults to 0 (Acoustic Grand Piano).
Only effective when `track` is None.
is_drum : bool
A boolean number that indicates whether it is a percussion track.
Defaults to False. Only effective when `track` is None.
name : str
The name of the track. Defaults to 'unknown'. Only effective when
`track` is None.
References
----------
[1] https://www.midi.org/specifications/item/gm-level-1-sound-set
"""
if track is not None:
if not isinstance(track, Track):
raise TypeError("`track` must be a pypianoroll.Track instance.")
track.check_validity()
else:
track = Track(pianoroll, program, is_drum, name)
self.tracks.append(track)
|
python
|
{
"resource": ""
}
|
q535
|
Multitrack.check_validity
|
train
|
def check_validity(self):
"""
Raise an error if any invalid attribute found.
Raises
------
TypeError
If an attribute has an invalid type.
ValueError
If an attribute has an invalid value (of the correct type).
"""
# tracks
for track in self.tracks:
if not isinstance(track, Track):
raise TypeError("`tracks` must be a list of "
"`pypianoroll.Track` instances.")
track.check_validity()
# tempo
if not isinstance(self.tempo, np.ndarray):
raise TypeError("`tempo` must be int or a numpy array.")
elif not np.issubdtype(self.tempo.dtype, np.number):
raise TypeError("Data type of `tempo` must be a subdtype of "
"np.number.")
elif self.tempo.ndim != 1:
raise ValueError("`tempo` must be a 1D numpy array.")
if np.any(self.tempo <= 0.0):
raise ValueError("`tempo` should contain only positive numbers.")
# downbeat
if self.downbeat is not None:
if not isinstance(self.downbeat, np.ndarray):
raise TypeError("`downbeat` must be a numpy array.")
if not np.issubdtype(self.downbeat.dtype, np.bool_):
raise TypeError("Data type of `downbeat` must be bool.")
if self.downbeat.ndim != 1:
raise ValueError("`downbeat` must be a 1D numpy array.")
# beat_resolution
if not isinstance(self.beat_resolution, int):
raise TypeError("`beat_resolution` must be int.")
if self.beat_resolution < 1:
raise ValueError("`beat_resolution` must be a positive integer.")
# name
if not isinstance(self.name, string_types):
raise TypeError("`name` must be a string.")
|
python
|
{
"resource": ""
}
|
q536
|
Multitrack.clip
|
train
|
def clip(self, lower=0, upper=127):
"""
Clip the pianorolls of all tracks by the given lower and upper bounds.
Parameters
----------
lower : int or float
The lower bound to clip the pianorolls. Defaults to 0.
upper : int or float
The upper bound to clip the pianorolls. Defaults to 127.
"""
for track in self.tracks:
track.clip(lower, upper)
|
python
|
{
"resource": ""
}
|
q537
|
Multitrack.get_downbeat_steps
|
train
|
def get_downbeat_steps(self):
"""
Return the indices of time steps that contain downbeats.
Returns
-------
downbeat_steps : list
The indices of time steps that contain downbeats.
"""
if self.downbeat is None:
return []
downbeat_steps = np.nonzero(self.downbeat)[0].tolist()
return downbeat_steps
|
python
|
{
"resource": ""
}
|
q538
|
Multitrack.get_empty_tracks
|
train
|
def get_empty_tracks(self):
"""
Return the indices of tracks with empty pianorolls.
Returns
-------
empty_track_indices : list
The indices of tracks with empty pianorolls.
"""
empty_track_indices = [idx for idx, track in enumerate(self.tracks)
if not np.any(track.pianoroll)]
return empty_track_indices
|
python
|
{
"resource": ""
}
|
q539
|
Multitrack.get_merged_pianoroll
|
train
|
def get_merged_pianoroll(self, mode='sum'):
"""
Return the merged pianoroll.
Parameters
----------
mode : {'sum', 'max', 'any'}
A string that indicates the merging strategy to apply along the
track axis. Default to 'sum'.
- In 'sum' mode, the merged pianoroll is the sum of all the
pianorolls. Note that for binarized pianorolls, integer summation
is performed.
- In 'max' mode, for each pixel, the maximum value among all the
pianorolls is assigned to the merged pianoroll.
- In 'any' mode, the value of a pixel in the merged pianoroll is
True if any of the pianorolls has nonzero value at that pixel;
False if all pianorolls are inactive (zero-valued) at that pixel.
Returns
-------
merged : np.ndarray, shape=(n_time_steps, 128)
The merged pianoroll.
"""
stacked = self.get_stacked_pianoroll()
if mode == 'any':
merged = np.any(stacked, axis=2)
elif mode == 'sum':
merged = np.sum(stacked, axis=2)
elif mode == 'max':
merged = np.max(stacked, axis=2)
else:
raise ValueError("`mode` must be one of {'max', 'sum', 'any'}.")
return merged
|
python
|
{
"resource": ""
}
|
q540
|
Multitrack.merge_tracks
|
train
|
def merge_tracks(self, track_indices=None, mode='sum', program=0,
is_drum=False, name='merged', remove_merged=False):
"""
Merge pianorolls of the tracks specified by `track_indices`. The merged
track will have program number as given by `program` and drum indicator
as given by `is_drum`. The merged track will be appended at the end of
the track list.
Parameters
----------
track_indices : list
The indices of tracks to be merged. Defaults to all the tracks.
mode : {'sum', 'max', 'any'}
A string that indicates the merging strategy to apply along the
track axis. Default to 'sum'.
- In 'sum' mode, the merged pianoroll is the sum of the collected
pianorolls. Note that for binarized pianorolls, integer summation
is performed.
- In 'max' mode, for each pixel, the maximum value among the
collected pianorolls is assigned to the merged pianoroll.
- In 'any' mode, the value of a pixel in the merged pianoroll is
True if any of the collected pianorolls has nonzero value at that
pixel; False if all the collected pianorolls are inactive
(zero-valued) at that pixel.
program: int
A program number according to General MIDI specification [1].
Available values are 0 to 127. Defaults to 0 (Acoustic Grand Piano).
is_drum : bool
A boolean number that indicates whether it is a percussion track.
Defaults to False.
name : str
A name to be assigned to the merged track. Defaults to 'merged'.
remove_merged : bool
True to remove the source tracks from the track list. False to keep
them. Defaults to False.
References
----------
[1] https://www.midi.org/specifications/item/gm-level-1-sound-set
"""
if mode not in ('max', 'sum', 'any'):
raise ValueError("`mode` must be one of {'max', 'sum', 'any'}.")
merged = self[track_indices].get_merged_pianoroll(mode)
merged_track = Track(merged, program, is_drum, name)
self.append_track(merged_track)
if remove_merged:
self.remove_tracks(track_indices)
|
python
|
{
"resource": ""
}
|
q541
|
Multitrack.pad_to_same
|
train
|
def pad_to_same(self):
"""Pad shorter pianorolls with zeros at the end along the time axis to
make the resulting pianoroll lengths the same as the maximum pianoroll
length among all the tracks."""
max_length = self.get_max_length()
for track in self.tracks:
if track.pianoroll.shape[0] < max_length:
track.pad(max_length - track.pianoroll.shape[0])
|
python
|
{
"resource": ""
}
|
q542
|
Multitrack.remove_tracks
|
train
|
def remove_tracks(self, track_indices):
"""
Remove tracks specified by `track_indices`.
Parameters
----------
track_indices : list
The indices of the tracks to be removed.
"""
if isinstance(track_indices, int):
track_indices = [track_indices]
self.tracks = [track for idx, track in enumerate(self.tracks)
if idx not in track_indices]
|
python
|
{
"resource": ""
}
|
q543
|
Multitrack.transpose
|
train
|
def transpose(self, semitone):
"""
Transpose the pianorolls of all tracks by a number of semitones, where
positive values are for higher key, while negative values are for lower
key. The drum tracks are ignored.
Parameters
----------
semitone : int
The number of semitones to transpose the pianorolls.
"""
for track in self.tracks:
if not track.is_drum:
track.transpose(semitone)
|
python
|
{
"resource": ""
}
|
q544
|
Multitrack.trim_trailing_silence
|
train
|
def trim_trailing_silence(self):
"""Trim the trailing silences of the pianorolls of all tracks. Trailing
silences are considered globally."""
active_length = self.get_active_length()
for track in self.tracks:
track.pianoroll = track.pianoroll[:active_length]
|
python
|
{
"resource": ""
}
|
q545
|
Multitrack.write
|
train
|
def write(self, filename):
"""
Write the multitrack pianoroll to a MIDI file.
Parameters
----------
filename : str
The name of the MIDI file to which the multitrack pianoroll is
written.
"""
if not filename.endswith(('.mid', '.midi', '.MID', '.MIDI')):
filename = filename + '.mid'
pm = self.to_pretty_midi()
pm.write(filename)
|
python
|
{
"resource": ""
}
|
q546
|
check_pianoroll
|
train
|
def check_pianoroll(arr):
"""
Return True if the array is a standard piano-roll matrix. Otherwise,
return False. Raise TypeError if the input object is not a numpy array.
"""
if not isinstance(arr, np.ndarray):
raise TypeError("`arr` must be of np.ndarray type")
if not (np.issubdtype(arr.dtype, np.bool_)
or np.issubdtype(arr.dtype, np.number)):
return False
if arr.ndim != 2:
return False
if arr.shape[1] != 128:
return False
return True
|
python
|
{
"resource": ""
}
|
q547
|
pad
|
train
|
def pad(obj, pad_length):
"""
Return a copy of the object with piano-roll padded with zeros at the end
along the time axis.
Parameters
----------
pad_length : int
The length to pad along the time axis with zeros.
"""
_check_supported(obj)
copied = deepcopy(obj)
copied.pad(pad_length)
return copied
|
python
|
{
"resource": ""
}
|
q548
|
pad_to_multiple
|
train
|
def pad_to_multiple(obj, factor):
"""
Return a copy of the object with its piano-roll padded with zeros at the
end along the time axis with the minimal length that make the length of
the resulting piano-roll a multiple of `factor`.
Parameters
----------
factor : int
The value which the length of the resulting piano-roll will be
a multiple of.
"""
_check_supported(obj)
copied = deepcopy(obj)
copied.pad_to_multiple(factor)
return copied
|
python
|
{
"resource": ""
}
|
q549
|
pad_to_same
|
train
|
def pad_to_same(obj):
"""
Return a copy of the object with shorter piano-rolls padded with zeros
at the end along the time axis to the length of the piano-roll with the
maximal length.
"""
if not isinstance(obj, Multitrack):
raise TypeError("Support only `pypianoroll.Multitrack` class objects")
copied = deepcopy(obj)
copied.pad_to_same()
return copied
|
python
|
{
"resource": ""
}
|
q550
|
save
|
train
|
def save(filepath, obj, compressed=True):
"""
Save the object to a .npz file.
Parameters
----------
filepath : str
The path to save the file.
obj: `pypianoroll.Multitrack` objects
The object to be saved.
"""
if not isinstance(obj, Multitrack):
raise TypeError("Support only `pypianoroll.Multitrack` class objects")
obj.save(filepath, compressed)
|
python
|
{
"resource": ""
}
|
q551
|
write
|
train
|
def write(obj, filepath):
"""
Write the object to a MIDI file.
Parameters
----------
filepath : str
The path to write the MIDI file.
"""
if not isinstance(obj, Multitrack):
raise TypeError("Support only `pypianoroll.Multitrack` class objects")
obj.write(filepath)
|
python
|
{
"resource": ""
}
|
q552
|
_validate_pianoroll
|
train
|
def _validate_pianoroll(pianoroll):
"""Raise an error if the input array is not a standard pianoroll."""
if not isinstance(pianoroll, np.ndarray):
raise TypeError("`pianoroll` must be of np.ndarray type.")
if not (np.issubdtype(pianoroll.dtype, np.bool_)
or np.issubdtype(pianoroll.dtype, np.number)):
raise TypeError("The data type of `pianoroll` must be np.bool_ or a "
"subdtype of np.number.")
if pianoroll.ndim != 2:
raise ValueError("`pianoroll` must have exactly two dimensions.")
if pianoroll.shape[1] != 128:
raise ValueError("The length of the second axis of `pianoroll` must be "
"128.")
|
python
|
{
"resource": ""
}
|
q553
|
_to_chroma
|
train
|
def _to_chroma(pianoroll):
"""Return the unnormalized chroma features of a pianoroll."""
_validate_pianoroll(pianoroll)
reshaped = pianoroll[:, :120].reshape(-1, 12, 10)
reshaped[..., :8] += pianoroll[:, 120:].reshape(-1, 1, 8)
return np.sum(reshaped, 1)
|
python
|
{
"resource": ""
}
|
q554
|
empty_beat_rate
|
train
|
def empty_beat_rate(pianoroll, beat_resolution):
"""Return the ratio of empty beats to the total number of beats in a
pianoroll."""
_validate_pianoroll(pianoroll)
reshaped = pianoroll.reshape(-1, beat_resolution * pianoroll.shape[1])
n_empty_beats = np.count_nonzero(reshaped.any(1))
return n_empty_beats / len(reshaped)
|
python
|
{
"resource": ""
}
|
q555
|
n_pitche_classes_used
|
train
|
def n_pitche_classes_used(pianoroll):
"""Return the number of unique pitch classes used in a pianoroll."""
_validate_pianoroll(pianoroll)
chroma = _to_chroma(pianoroll)
return np.count_nonzero(np.any(chroma, 0))
|
python
|
{
"resource": ""
}
|
q556
|
polyphonic_rate
|
train
|
def polyphonic_rate(pianoroll, threshold=2):
"""Return the ratio of the number of time steps where the number of pitches
being played is larger than `threshold` to the total number of time steps
in a pianoroll."""
_validate_pianoroll(pianoroll)
n_poly = np.count_nonzero(np.count_nonzero(pianoroll, 1) > threshold)
return n_poly / len(pianoroll)
|
python
|
{
"resource": ""
}
|
q557
|
in_scale_rate
|
train
|
def in_scale_rate(pianoroll, key=3, kind='major'):
"""Return the ratio of the number of nonzero entries that lie in a specific
scale to the total number of nonzero entries in a pianoroll. Default to C
major scale."""
if not isinstance(key, int):
raise TypeError("`key` must an integer.")
if key > 11 or key < 0:
raise ValueError("`key` must be in an integer in between 0 and 11.")
if kind not in ('major', 'minor'):
raise ValueError("`kind` must be one of 'major' or 'minor'.")
_validate_pianoroll(pianoroll)
def _scale_mask(key, kind):
"""Return a scale mask for the given key. Default to C major scale."""
if kind == 'major':
a_scale_mask = np.array([0, 1, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1], bool)
else:
a_scale_mask = np.array([1, 0, 1, 0, 1, 1, 0, 1, 0, 1, 0, 1], bool)
return np.roll(a_scale_mask, key)
chroma = _to_chroma(pianoroll)
scale_mask = _scale_mask(key, kind)
n_in_scale = np.sum(scale_mask.reshape(-1, 12) * chroma)
return n_in_scale / np.count_nonzero(pianoroll)
|
python
|
{
"resource": ""
}
|
q558
|
Track.assign_constant
|
train
|
def assign_constant(self, value, dtype=None):
"""
Assign a constant value to all nonzeros in the pianoroll. If the
pianoroll is not binarized, its data type will be preserved. If the
pianoroll is binarized, it will be casted to the type of `value`.
Arguments
---------
value : int or float
The constant value to be assigned to all the nonzeros in the
pianoroll.
"""
if not self.is_binarized():
self.pianoroll[self.pianoroll.nonzero()] = value
return
if dtype is None:
if isinstance(value, int):
dtype = int
elif isinstance(value, float):
dtype = float
nonzero = self.pianoroll.nonzero()
self.pianoroll = np.zeros(self.pianoroll.shape, dtype)
self.pianoroll[nonzero] = value
|
python
|
{
"resource": ""
}
|
q559
|
Track.binarize
|
train
|
def binarize(self, threshold=0):
"""
Binarize the pianoroll.
Parameters
----------
threshold : int or float
A threshold used to binarize the pianorolls. Defaults to zero.
"""
if not self.is_binarized():
self.pianoroll = (self.pianoroll > threshold)
|
python
|
{
"resource": ""
}
|
q560
|
Track.check_validity
|
train
|
def check_validity(self):
""""Raise error if any invalid attribute found."""
# pianoroll
if not isinstance(self.pianoroll, np.ndarray):
raise TypeError("`pianoroll` must be a numpy array.")
if not (np.issubdtype(self.pianoroll.dtype, np.bool_)
or np.issubdtype(self.pianoroll.dtype, np.number)):
raise TypeError("The data type of `pianoroll` must be np.bool_ or "
"a subdtype of np.number.")
if self.pianoroll.ndim != 2:
raise ValueError("`pianoroll` must have exactly two dimensions.")
if self.pianoroll.shape[1] != 128:
raise ValueError("The length of the second axis of `pianoroll` "
"must be 128.")
# program
if not isinstance(self.program, int):
raise TypeError("`program` must be int.")
if self.program < 0 or self.program > 127:
raise ValueError("`program` must be in between 0 to 127.")
# is_drum
if not isinstance(self.is_drum, bool):
raise TypeError("`is_drum` must be bool.")
# name
if not isinstance(self.name, string_types):
raise TypeError("`name` must be a string.")
|
python
|
{
"resource": ""
}
|
q561
|
Track.clip
|
train
|
def clip(self, lower=0, upper=127):
"""
Clip the pianoroll by the given lower and upper bounds.
Parameters
----------
lower : int or float
The lower bound to clip the pianoroll. Defaults to 0.
upper : int or float
The upper bound to clip the pianoroll. Defaults to 127.
"""
self.pianoroll = self.pianoroll.clip(lower, upper)
|
python
|
{
"resource": ""
}
|
q562
|
Track.is_binarized
|
train
|
def is_binarized(self):
"""
Return True if the pianoroll is already binarized. Otherwise, return
False.
Returns
-------
is_binarized : bool
True if the pianoroll is already binarized; otherwise, False.
"""
is_binarized = np.issubdtype(self.pianoroll.dtype, np.bool_)
return is_binarized
|
python
|
{
"resource": ""
}
|
q563
|
Track.pad
|
train
|
def pad(self, pad_length):
"""
Pad the pianoroll with zeros at the end along the time axis.
Parameters
----------
pad_length : int
The length to pad with zeros along the time axis.
"""
self.pianoroll = np.pad(
self.pianoroll, ((0, pad_length), (0, 0)), 'constant')
|
python
|
{
"resource": ""
}
|
q564
|
Track.pad_to_multiple
|
train
|
def pad_to_multiple(self, factor):
"""
Pad the pianoroll with zeros at the end along the time axis with the
minimum length that makes the resulting pianoroll length a multiple of
`factor`.
Parameters
----------
factor : int
The value which the length of the resulting pianoroll will be
a multiple of.
"""
remainder = self.pianoroll.shape[0] % factor
if remainder:
pad_width = ((0, (factor - remainder)), (0, 0))
self.pianoroll = np.pad(self.pianoroll, pad_width, 'constant')
|
python
|
{
"resource": ""
}
|
q565
|
Track.transpose
|
train
|
def transpose(self, semitone):
"""
Transpose the pianoroll by a number of semitones, where positive
values are for higher key, while negative values are for lower key.
Parameters
----------
semitone : int
The number of semitones to transpose the pianoroll.
"""
if semitone > 0 and semitone < 128:
self.pianoroll[:, semitone:] = self.pianoroll[:, :(128 - semitone)]
self.pianoroll[:, :semitone] = 0
elif semitone < 0 and semitone > -128:
self.pianoroll[:, :(128 + semitone)] = self.pianoroll[:, -semitone:]
self.pianoroll[:, (128 + semitone):] = 0
|
python
|
{
"resource": ""
}
|
q566
|
Track.trim_trailing_silence
|
train
|
def trim_trailing_silence(self):
"""Trim the trailing silence of the pianoroll."""
length = self.get_active_length()
self.pianoroll = self.pianoroll[:length]
|
python
|
{
"resource": ""
}
|
q567
|
plot_conv_weights
|
train
|
def plot_conv_weights(layer, figsize=(6, 6)):
"""Plot the weights of a specific layer.
Only really makes sense with convolutional layers.
Parameters
----------
layer : lasagne.layers.Layer
"""
W = layer.W.get_value()
shape = W.shape
nrows = np.ceil(np.sqrt(shape[0])).astype(int)
ncols = nrows
for feature_map in range(shape[1]):
figs, axes = plt.subplots(nrows, ncols, figsize=figsize, squeeze=False)
for ax in axes.flatten():
ax.set_xticks([])
ax.set_yticks([])
ax.axis('off')
for i, (r, c) in enumerate(product(range(nrows), range(ncols))):
if i >= shape[0]:
break
axes[r, c].imshow(W[i, feature_map], cmap='gray',
interpolation='none')
return plt
|
python
|
{
"resource": ""
}
|
q568
|
plot_conv_activity
|
train
|
def plot_conv_activity(layer, x, figsize=(6, 8)):
"""Plot the acitivities of a specific layer.
Only really makes sense with layers that work 2D data (2D
convolutional layers, 2D pooling layers ...).
Parameters
----------
layer : lasagne.layers.Layer
x : numpy.ndarray
Only takes one sample at a time, i.e. x.shape[0] == 1.
"""
if x.shape[0] != 1:
raise ValueError("Only one sample can be plotted at a time.")
# compile theano function
xs = T.tensor4('xs').astype(theano.config.floatX)
get_activity = theano.function([xs], get_output(layer, xs))
activity = get_activity(x)
shape = activity.shape
nrows = np.ceil(np.sqrt(shape[1])).astype(int)
ncols = nrows
figs, axes = plt.subplots(nrows + 1, ncols, figsize=figsize, squeeze=False)
axes[0, ncols // 2].imshow(1 - x[0][0], cmap='gray',
interpolation='none')
axes[0, ncols // 2].set_title('original')
for ax in axes.flatten():
ax.set_xticks([])
ax.set_yticks([])
ax.axis('off')
for i, (r, c) in enumerate(product(range(nrows), range(ncols))):
if i >= shape[1]:
break
ndim = activity[0][i].ndim
if ndim != 2:
raise ValueError("Wrong number of dimensions, image data should "
"have 2, instead got {}".format(ndim))
axes[r + 1, c].imshow(-activity[0][i], cmap='gray',
interpolation='none')
return plt
|
python
|
{
"resource": ""
}
|
q569
|
occlusion_heatmap
|
train
|
def occlusion_heatmap(net, x, target, square_length=7):
"""An occlusion test that checks an image for its critical parts.
In this function, a square part of the image is occluded (i.e. set
to 0) and then the net is tested for its propensity to predict the
correct label. One should expect that this propensity shrinks of
critical parts of the image are occluded. If not, this indicates
overfitting.
Depending on the depth of the net and the size of the image, this
function may take awhile to finish, since one prediction for each
pixel of the image is made.
Currently, all color channels are occluded at the same time. Also,
this does not really work if images are randomly distorted by the
batch iterator.
See paper: Zeiler, Fergus 2013
Parameters
----------
net : NeuralNet instance
The neural net to test.
x : np.array
The input data, should be of shape (1, c, x, y). Only makes
sense with image data.
target : int
The true value of the image. If the net makes several
predictions, say 10 classes, this indicates which one to look
at.
square_length : int (default=7)
The length of the side of the square that occludes the image.
Must be an odd number.
Results
-------
heat_array : np.array (with same size as image)
An 2D np.array that at each point (i, j) contains the predicted
probability of the correct class if the image is occluded by a
square with center (i, j).
"""
if (x.ndim != 4) or x.shape[0] != 1:
raise ValueError("This function requires the input data to be of "
"shape (1, c, x, y), instead got {}".format(x.shape))
if square_length % 2 == 0:
raise ValueError("Square length has to be an odd number, instead "
"got {}.".format(square_length))
num_classes = get_output_shape(net.layers_[-1])[1]
img = x[0].copy()
bs, col, s0, s1 = x.shape
heat_array = np.zeros((s0, s1))
pad = square_length // 2 + 1
x_occluded = np.zeros((s1, col, s0, s1), dtype=img.dtype)
probs = np.zeros((s0, s1, num_classes))
# generate occluded images
for i in range(s0):
# batch s1 occluded images for faster prediction
for j in range(s1):
x_pad = np.pad(img, ((0, 0), (pad, pad), (pad, pad)), 'constant')
x_pad[:, i:i + square_length, j:j + square_length] = 0.
x_occluded[j] = x_pad[:, pad:-pad, pad:-pad]
y_proba = net.predict_proba(x_occluded)
probs[i] = y_proba.reshape(s1, num_classes)
# from predicted probabilities, pick only those of target class
for i in range(s0):
for j in range(s1):
heat_array[i, j] = probs[i, j, target]
return heat_array
|
python
|
{
"resource": ""
}
|
q570
|
plot_occlusion
|
train
|
def plot_occlusion(net, X, target, square_length=7, figsize=(9, None)):
"""Plot which parts of an image are particularly import for the
net to classify the image correctly.
See paper: Zeiler, Fergus 2013
Parameters
----------
net : NeuralNet instance
The neural net to test.
X : numpy.array
The input data, should be of shape (b, c, 0, 1). Only makes
sense with image data.
target : list or numpy.array of ints
The true values of the image. If the net makes several
predictions, say 10 classes, this indicates which one to look
at. If more than one sample is passed to X, each of them needs
its own target.
square_length : int (default=7)
The length of the side of the square that occludes the image.
Must be an odd number.
figsize : tuple (int, int)
Size of the figure.
Plots
-----
Figure with 3 subplots: the original image, the occlusion heatmap,
and both images super-imposed.
"""
return _plot_heat_map(
net, X, figsize, lambda net, X, n: occlusion_heatmap(
net, X, target[n], square_length))
|
python
|
{
"resource": ""
}
|
q571
|
multiclass_logloss
|
train
|
def multiclass_logloss(actual, predicted, eps=1e-15):
"""Multi class version of Logarithmic Loss metric.
:param actual: Array containing the actual target classes
:param predicted: Matrix with class predictions, one probability per class
"""
# Convert 'actual' to a binary array if it's not already:
if len(actual.shape) == 1:
actual2 = np.zeros((actual.shape[0], predicted.shape[1]))
for i, val in enumerate(actual):
actual2[i, val] = 1
actual = actual2
clip = np.clip(predicted, eps, 1 - eps)
rows = actual.shape[0]
vsota = np.sum(actual * np.log(clip))
return -1.0 / rows * vsota
|
python
|
{
"resource": ""
}
|
q572
|
objective
|
train
|
def objective(layers,
loss_function,
target,
aggregate=aggregate,
deterministic=False,
l1=0,
l2=0,
get_output_kw=None):
"""
Default implementation of the NeuralNet objective.
:param layers: The underlying layers of the NeuralNetwork
:param loss_function: The callable loss function to use
:param target: the expected output
:param aggregate: the aggregation function to use
:param deterministic: Whether or not to get a deterministic output
:param l1: Optional l1 regularization parameter
:param l2: Optional l2 regularization parameter
:param get_output_kw: optional kwargs to pass to
:meth:`NeuralNetwork.get_output`
:return: The total calculated loss
"""
if get_output_kw is None:
get_output_kw = {}
output_layer = layers[-1]
network_output = get_output(
output_layer, deterministic=deterministic, **get_output_kw)
loss = aggregate(loss_function(network_output, target))
if l1:
loss += regularization.regularize_layer_params(
layers.values(), regularization.l1) * l1
if l2:
loss += regularization.regularize_layer_params(
layers.values(), regularization.l2) * l2
return loss
|
python
|
{
"resource": ""
}
|
q573
|
NeuralNet.initialize
|
train
|
def initialize(self):
"""Initializes the network. Checks that no extra kwargs were
passed to the constructor, and compiles the train, predict,
and evaluation functions.
Subsequent calls to this function will return without any action.
"""
if getattr(self, '_initialized', False):
return
out = getattr(self, '_output_layers', None)
if out is None:
self.initialize_layers()
self._check_for_unused_kwargs()
iter_funcs = self._create_iter_funcs(
self.layers_, self.objective, self.update,
self.y_tensor_type,
)
self.train_iter_, self.eval_iter_, self.predict_iter_ = iter_funcs
self._initialized = True
|
python
|
{
"resource": ""
}
|
q574
|
NeuralNet.fit
|
train
|
def fit(self, X, y, epochs=None):
"""
Runs the training loop for a given number of epochs
:param X: The input data
:param y: The ground truth
:param epochs: The number of epochs to run, if `None` runs for the
network's :attr:`max_epochs`
:return: This instance
"""
if self.check_input:
X, y = self._check_good_input(X, y)
if self.use_label_encoder:
self.enc_ = LabelEncoder()
y = self.enc_.fit_transform(y).astype(np.int32)
self.classes_ = self.enc_.classes_
self.initialize()
try:
self.train_loop(X, y, epochs=epochs)
except KeyboardInterrupt:
pass
return self
|
python
|
{
"resource": ""
}
|
q575
|
NeuralNet.partial_fit
|
train
|
def partial_fit(self, X, y, classes=None):
"""
Runs a single epoch using the provided data
:return: This instance
"""
return self.fit(X, y, epochs=1)
|
python
|
{
"resource": ""
}
|
q576
|
ByDateQuerySetMixin.narrow
|
train
|
def narrow(self, **kwargs):
"""Up-to including"""
from_date = kwargs.pop('from_date', None)
to_date = kwargs.pop('to_date', None)
date = kwargs.pop('date', None)
qs = self
if from_date:
qs = qs.filter(date__gte=from_date)
if to_date:
qs = qs.filter(date__lte=to_date)
if date:
qs = qs.filter(date=date)
return super(ByDateQuerySetMixin, qs).narrow(**kwargs)
|
python
|
{
"resource": ""
}
|
q577
|
set_environment_variables
|
train
|
def set_environment_variables(json_file_path):
"""
Read and set environment variables from a flat json file.
Bear in mind that env vars set this way and later on read using
`os.getenv` function will be strings since after all env vars are just
that - plain strings.
Json file example:
```
{
"FOO": "bar",
"BAZ": true
}
```
:param json_file_path: path to flat json file
:type json_file_path: str
"""
if json_file_path:
with open(json_file_path) as json_file:
env_vars = json.loads(json_file.read())
export_variables(env_vars)
|
python
|
{
"resource": ""
}
|
q578
|
millis_interval
|
train
|
def millis_interval(start, end):
"""start and end are datetime instances"""
diff = end - start
millis = diff.days * 24 * 60 * 60 * 1000
millis += diff.seconds * 1000
millis += diff.microseconds / 1000
return millis
|
python
|
{
"resource": ""
}
|
q579
|
Script._import_lua
|
train
|
def _import_lua(load_dependencies=True):
"""
Import lua and dependencies.
:param load_dependencies: should Lua library dependencies be loaded?
:raises: RuntimeError if Lua is not available
"""
try:
import lua
except ImportError:
raise RuntimeError("Lua not installed")
lua_globals = lua.globals()
if load_dependencies:
Script._import_lua_dependencies(lua, lua_globals)
return lua, lua_globals
|
python
|
{
"resource": ""
}
|
q580
|
Script._import_lua_dependencies
|
train
|
def _import_lua_dependencies(lua, lua_globals):
"""
Imports lua dependencies that are supported by redis lua scripts.
The current implementation is fragile to the target platform and lua version
and may be disabled if these imports are not needed.
Included:
- cjson lib.
Pending:
- base lib.
- table lib.
- string lib.
- math lib.
- debug lib.
- cmsgpack lib.
"""
if sys.platform not in ('darwin', 'windows'):
import ctypes
ctypes.CDLL('liblua5.2.so', mode=ctypes.RTLD_GLOBAL)
try:
lua_globals.cjson = lua.eval('require "cjson"')
except RuntimeError:
raise RuntimeError("cjson not installed")
|
python
|
{
"resource": ""
}
|
q581
|
MockRedis.lock
|
train
|
def lock(self, key, timeout=0, sleep=0):
"""Emulate lock."""
return MockRedisLock(self, key, timeout, sleep)
|
python
|
{
"resource": ""
}
|
q582
|
MockRedis.keys
|
train
|
def keys(self, pattern='*'):
"""Emulate keys."""
# making sure the pattern is unicode/str.
try:
pattern = pattern.decode('utf-8')
# This throws an AttributeError in python 3, or an
# UnicodeEncodeError in python 2
except (AttributeError, UnicodeEncodeError):
pass
# Make regex out of glob styled pattern.
regex = fnmatch.translate(pattern)
regex = re.compile(re.sub(r'(^|[^\\])\.', r'\1[^/]', regex))
# Find every key that matches the pattern
return [key for key in self.redis.keys() if regex.match(key.decode('utf-8'))]
|
python
|
{
"resource": ""
}
|
q583
|
MockRedis.delete
|
train
|
def delete(self, *keys):
"""Emulate delete."""
key_counter = 0
for key in map(self._encode, keys):
if key in self.redis:
del self.redis[key]
key_counter += 1
if key in self.timeouts:
del self.timeouts[key]
return key_counter
|
python
|
{
"resource": ""
}
|
q584
|
MockRedis.do_expire
|
train
|
def do_expire(self):
"""
Expire objects assuming now == time
"""
# Deep copy to avoid RuntimeError: dictionary changed size during iteration
_timeouts = deepcopy(self.timeouts)
for key, value in _timeouts.items():
if value - self.clock.now() < timedelta(0):
del self.timeouts[key]
# removing the expired key
if key in self.redis:
self.redis.pop(key, None)
|
python
|
{
"resource": ""
}
|
q585
|
MockRedis.set
|
train
|
def set(self, key, value, ex=None, px=None, nx=False, xx=False):
"""
Set the ``value`` for the ``key`` in the context of the provided kwargs.
As per the behavior of the redis-py lib:
If nx and xx are both set, the function does nothing and None is returned.
If px and ex are both set, the preference is given to px.
If the key is not set for some reason, the lib function returns None.
"""
key = self._encode(key)
value = self._encode(value)
if nx and xx:
return None
mode = "nx" if nx else "xx" if xx else None
if self._should_set(key, mode):
expire = None
if ex is not None:
expire = ex if isinstance(ex, timedelta) else timedelta(seconds=ex)
if px is not None:
expire = px if isinstance(px, timedelta) else timedelta(milliseconds=px)
if expire is not None and expire.total_seconds() <= 0:
raise ResponseError("invalid expire time in SETEX")
result = self._set(key, value)
if expire:
self._expire(key, expire)
return result
|
python
|
{
"resource": ""
}
|
q586
|
MockRedis._should_set
|
train
|
def _should_set(self, key, mode):
"""
Determine if it is okay to set a key.
If the mode is None, returns True, otherwise, returns True of false based on
the value of ``key`` and the ``mode`` (nx | xx).
"""
if mode is None or mode not in ["nx", "xx"]:
return True
if mode == "nx":
if key in self.redis:
# nx means set only if key is absent
# false if the key already exists
return False
elif key not in self.redis:
# at this point mode can only be xx
# xx means set only if the key already exists
# false if is absent
return False
# for all other cases, return true
return True
|
python
|
{
"resource": ""
}
|
q587
|
MockRedis.setex
|
train
|
def setex(self, name, time, value):
"""
Set the value of ``name`` to ``value`` that expires in ``time``
seconds. ``time`` can be represented by an integer or a Python
timedelta object.
"""
if not self.strict:
# when not strict mode swap value and time args order
time, value = value, time
return self.set(name, value, ex=time)
|
python
|
{
"resource": ""
}
|
q588
|
MockRedis.psetex
|
train
|
def psetex(self, key, time, value):
"""
Set the value of ``key`` to ``value`` that expires in ``time``
milliseconds. ``time`` can be represented by an integer or a Python
timedelta object.
"""
return self.set(key, value, px=time)
|
python
|
{
"resource": ""
}
|
q589
|
MockRedis.setnx
|
train
|
def setnx(self, key, value):
"""Set the value of ``key`` to ``value`` if key doesn't exist"""
return self.set(key, value, nx=True)
|
python
|
{
"resource": ""
}
|
q590
|
MockRedis.setbit
|
train
|
def setbit(self, key, offset, value):
"""
Set the bit at ``offset`` in ``key`` to ``value``.
"""
key = self._encode(key)
index, bits, mask = self._get_bits_and_offset(key, offset)
if index >= len(bits):
bits.extend(b"\x00" * (index + 1 - len(bits)))
prev_val = 1 if (bits[index] & mask) else 0
if value:
bits[index] |= mask
else:
bits[index] &= ~mask
self.redis[key] = bytes(bits)
return prev_val
|
python
|
{
"resource": ""
}
|
q591
|
MockRedis.getbit
|
train
|
def getbit(self, key, offset):
"""
Returns the bit value at ``offset`` in ``key``.
"""
key = self._encode(key)
index, bits, mask = self._get_bits_and_offset(key, offset)
if index >= len(bits):
return 0
return 1 if (bits[index] & mask) else 0
|
python
|
{
"resource": ""
}
|
q592
|
MockRedis.hexists
|
train
|
def hexists(self, hashkey, attribute):
"""Emulate hexists."""
redis_hash = self._get_hash(hashkey, 'HEXISTS')
return self._encode(attribute) in redis_hash
|
python
|
{
"resource": ""
}
|
q593
|
MockRedis.hget
|
train
|
def hget(self, hashkey, attribute):
"""Emulate hget."""
redis_hash = self._get_hash(hashkey, 'HGET')
return redis_hash.get(self._encode(attribute))
|
python
|
{
"resource": ""
}
|
q594
|
MockRedis.hmset
|
train
|
def hmset(self, hashkey, value):
"""Emulate hmset."""
redis_hash = self._get_hash(hashkey, 'HMSET', create=True)
for key, value in value.items():
attribute = self._encode(key)
redis_hash[attribute] = self._encode(value)
return True
|
python
|
{
"resource": ""
}
|
q595
|
MockRedis.hmget
|
train
|
def hmget(self, hashkey, keys, *args):
"""Emulate hmget."""
redis_hash = self._get_hash(hashkey, 'HMGET')
attributes = self._list_or_args(keys, args)
return [redis_hash.get(self._encode(attribute)) for attribute in attributes]
|
python
|
{
"resource": ""
}
|
q596
|
MockRedis.hset
|
train
|
def hset(self, hashkey, attribute, value):
"""Emulate hset."""
redis_hash = self._get_hash(hashkey, 'HSET', create=True)
attribute = self._encode(attribute)
attribute_present = attribute in redis_hash
redis_hash[attribute] = self._encode(value)
return long(0) if attribute_present else long(1)
|
python
|
{
"resource": ""
}
|
q597
|
MockRedis.hsetnx
|
train
|
def hsetnx(self, hashkey, attribute, value):
"""Emulate hsetnx."""
redis_hash = self._get_hash(hashkey, 'HSETNX', create=True)
attribute = self._encode(attribute)
if attribute in redis_hash:
return long(0)
else:
redis_hash[attribute] = self._encode(value)
return long(1)
|
python
|
{
"resource": ""
}
|
q598
|
MockRedis.hincrby
|
train
|
def hincrby(self, hashkey, attribute, increment=1):
"""Emulate hincrby."""
return self._hincrby(hashkey, attribute, 'HINCRBY', long, increment)
|
python
|
{
"resource": ""
}
|
q599
|
MockRedis.hincrbyfloat
|
train
|
def hincrbyfloat(self, hashkey, attribute, increment=1.0):
"""Emulate hincrbyfloat."""
return self._hincrby(hashkey, attribute, 'HINCRBYFLOAT', float, increment)
|
python
|
{
"resource": ""
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.