Package pycv :: Package cs :: Package stats :: Module stats :: Class BasicStats

Class BasicStats

Known Subclasses:: Stats2, Stats2e, Stats2m


A few statistics of a random tensor (RT) are packed as a tuple.
This class represents the collection of some statistics of J RTs.
The class contains a few variables:
    J: the number of RTs
    shape: the shape represents of how the dimensions of an RT are organized
    d: the number of dimensions of an RT
    orders: an array of statistic orders
    A: the numpy.array (tensor) representing the whole data

Instance Methods

[hide private]

__init__(self, J, shape, orders, A=None)
Initialize BasicStats with J, shape, orders, and optionally A.

source code

get_indices(self, i)
Get the indices of the i-th type of statistics.

source code

get_stat(self, i)
Get the tensor of statistics of the i-th type.

source code

get_cond_stat(self, j, i)
Get the class-conditional tensor of statistics of the ith-type.

source code

set_cond_stat(self, j, i, A)
Set the class-conditional tensor of statistics of the ith-type.

source code

_new_dA(self, new_shape, new_A)
This function is to be virtuallized and used internally only.

source code

take(self, indices)
Flatten the shape of the RTs, then take only dimensions indexed in 'indices' and remove the remaining dimensions.

source code

Method Details

[hide private]

get_indices(self, i)

source code

Get the indices of the i-th type of statistics.

Input:: i: the index
Output:: (istart,iend): the indices

get_stat(self, i)

source code

Get the tensor of statistics of the i-th type.

Input:: i: the index
Output:: A: the tensor of statistics of the i-th type with shape (J,)+shape^rank[i]

get_cond_stat(self, j, i)

source code

Get the class-conditional tensor of statistics of the ith-type.

Input:

j: the index i: the class

Output:

A: the class-conditional tensor of statistics of the i-th type: of shape shape^rank[i]

set_cond_stat(self, j, i, A)

source code

Set the class-conditional tensor of statistics of the ith-type.

Input:
    j: the index
    i: the class
    A: the class-conditional tensor of statistics of the i-th type
        of shape shape^rank[i]

_new_dA(self, new_shape, new_A)

source code

This function is to be virtuallized and used internally only. To create a copy of this 'self' with new 'shape' and new 'A'.

take(self, indices)

source code

Flatten the shape of the RTs, then take only dimensions indexed in 'indices' and remove the remaining dimensions.

Input:: indices: a numpy.array of dimensions to be kept.
Output:: B: the BasicStats obtained from the remaining dimensions.