def normalize_sample(x, group=[], mode='PercentSignalChange', baseline='All',
zero_threshold=1, verbose=True):
"""
Apply normalization
Parameters
----------
x : array
Input data array (sample num * feature num)
group : array_like
Group vector (length = sample num)
Mode : str
Normalization mode ('PercentSignalChange', 'Zscore', 'DivideMean', or 'SubtractMean'; default = 'PercentSignalChange')
Baseline : array_like or str
Baseline index vector (default: 'All')
ZeroThreshold : float
Zero threshold (default: 1)
Returns
-------
y : array
Normalized data array (sample num * feature num)
"""
if verbose:
print_start_msg()
p = Normalize()
y, _ = p.run(x, group, mode = mode, baseline = baseline, zero_threshold = zero_threshold)
if verbose:
print_finish_msg()
return y
def detrend_sample(x, group=[], keep_mean=True, verbose=True):
"""
Apply linear detrend
Parameters
----------
x : array
Input data array (sample num * feature num)
group : array_like
Group vector (length = sample num)
Returns
-------
y : array
Detrended data array (group num * feature num)
"""
if verbose:
print_start_msg()
p = Detrender()
y, _ = p.run(x, group, keep_mean=keep_mean)
if verbose:
print_finish_msg()
return y
def regressout(x, group=[], regressor=[], remove_dc=True, linear_detrend=True, verbose=True):
'''Remove nuisance regressors.
Parameters
----------
x : array, shape = (n_sample, n_feature)
Input data array
group : array_like, lenght = n_sample
Group vector.
regressor : array_like, shape = (n_sample, n_regressor)
Nuisance regressors.
remove_dc : bool
Remove DC component (signal mean) or not (default: True).
linear_detrend : bool
Remove linear trend or not (default: True).
Returns
-------
y : array, shape = (n_sample, n_feature)
Signal without nuisance regressors.
'''
if verbose:
print_start_msg()
p = Regressout()
y, _ = p.run(x, group, regressor=regressor, remove_dc=remove_dc, linear_detrend=linear_detrend)
if verbose:
print_finish_msg()
return y
def average_sample(x, group=[], verbose=True):
"""
Average samples within groups
Parameters
----------
x : array
Input data array (sample num * feature num)
group : array_like
Group vector (length = sample num)
Returns
-------
y : array
Averaged data array (group num * feature num)
index_map : array_like
Vector mapping row indexes from y to x (length = group num)
"""
if verbose:
print_start_msg()
p = Average()
y, ind_map = p.run(x, group)
if verbose:
print_finish_msg()
return y, ind_map
def shift_sample(x, group=[], shift_size = 1, verbose = True):
"""
Shift sample within groups
Parameters
----------
x : array
Input data (sample num * feature num)
group : array_like
Group vector (length: sample num)
shift_size : int
Shift size (default: 1)
Returns
-------
y : array
Averaged data array (group num * feature num)
index_map : array_like
Vector mapping row indexes from y to x (length: group num)
Example
-------
import numpy as np
from bdpy.preprocessor import shift_sample
x = np.array([[ 1, 2, 3 ],
[ 11, 12, 13 ],
[ 21, 22, 23 ],
[ 31, 32, 33 ],
[ 41, 42, 43 ],
[ 51, 52, 53 ]])
grp = np.array([ 1, 1, 1, 2, 2, 2 ])
shift_size = 1
y, index = shift_sample(x, grp, shift_size)
# >>> y
# array([[11, 12, 13],
# [21, 22, 23],
# [41, 42, 43],
# [51, 52, 53]])
# >>> index
# array([0, 1, 3, 4])
"""
if verbose:
print_start_msg()
p = ShiftSample()
y, index_map = p.run(x, group, shift_size = shift_size)
if verbose:
print_finish_msg()
return y, index_map
def select_top(data, value, num, axis=0, verbose=True):
"""
Select top `num` features of `value` from `data`
Parameters
----------
data : array
Data matrix
value : array_like
Vector of values
num : int
Number of selected features
Returns
-------
selected_data : array
Selected data matrix
selected_index : array
Index of selected data
"""
if verbose:
print_start_msg()
num_elem = data.shape[axis]
sorted_index = np.argsort(value)[::-1]
rank = np.zeros(num_elem, dtype=np.int)
rank[sorted_index] = np.array(range(0, num_elem))
selected_index_bool = rank < num
if axis == 0:
selected_data = data[selected_index_bool, :]
selected_index = np.array(range(0, num_elem),
dtype=np.int)[selected_index_bool]
elif axis == 1:
selected_data = data[:, selected_index_bool]
selected_index = np.array(range(0, num_elem),
dtype=np.int)[selected_index_bool]
else:
raise ValueError('Invalid axis')
if verbose:
print_finish_msg()
return selected_data, selected_index
def reduce_outlier(x, group=[], std=True, maxmin=True, remove=False, dimension=1, n_iter=10, std_threshold=3, max_value=None, min_value=None, verbose=True):
'''Outlier reduction.'''
if verbose:
print_start_msg()
if remove:
raise NotImplementedError('"remove" option is not implemented yet.')
p = ReduceOutlier()
y, _ = p.run(x, group, std=std, maxmin=maxmin, dimension=dimension, n_iter=n_iter, std_threshold=std_threshold, max_value=max_value, min_value=min_value)
if verbose:
print_finish_msg()
return y
