dasf.feature_extraction

Submodules

Classes

`Histogram`	Operator to extract the histogram of a data.
`ConcatenateToArray`	Concatenate data from different Arrays into a single array.
`GetSubDataframe`	Get the first x% samples from the dataset.
`SampleDataframe`	Return a subset with random samples of the original dataset.

Package Contents

class dasf.feature_extraction.Histogram(bins=None, range=None, normed=False, weights=None, density=None, *args, **kwargs)[source]

Bases: dasf.transforms.base.TargeteredTransform, dasf.transforms.base.Transform

Operator to extract the histogram of a data.

Parameters

binsOptional[int]: Number of bins (the default is None).
rangetuple: 2-element tuple with the lower and upper range of the bins. If not provided, range is simply (X.min(), X.max()) (the default is None).
normedbool: If the historgram must be normalized (the default is False).
weightstype: An array of weights, of the same shape as X. Each value in a only contributes its associated weight towards the bin count (the default is None).
densitytype: If False, the result will contain the number of samples in each bin. If True, the result is the value of the probability density function at the bin, normalized such that the integral over the range is 1 (the default is None).

Attributes

bins range normed weights density

Generic constructor of the class Histogram.

_bins

_range

_normed

_weights

_density

_lazy_transform_generic(X)[source]

Compute the histogram of a dataset using Dask.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

_transform_generic(X, xp)[source]

Compute the histogram of a dataset using local libraries.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

_lazy_transform_cpu(X)[source]

Compute the histogram of a dataset using Dask with CPUs only.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

_lazy_transform_gpu(X, **kwargs)[source]

Compute the histogram of a dataset using Dask with GPUs only.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

_transform_cpu(X, **kwargs)[source]

Compute the histogram of a dataset using CPU only.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

_transform_gpu(X, **kwargs)[source]

Compute the histogram of a dataset using GPU only.

Parameters

Xarray_like: Input data. The histogram is computed over the flattened array.

Returns

histarray: The values of the histogram. See density and weights for a description of the possible semantics. If weights are given, hist.dtype will be taken from weights.
bin_edgesarray of dtype float: Return the bin edges (length(hist)+1).

Parameters:

bins (int)
range (tuple)
normed (bool)

class dasf.feature_extraction.ConcatenateToArray(flatten=False)[source]

Bases: dasf.transforms.base.Transform

Concatenate data from different Arrays into a single array.

Parameters

flattenbool: If the arrays must be flatten prior concatenating. If False, the arrays must share the shape of last dimansions in order to be concatenated (the default is False).

flatten

__transform_generic(xp, **kwargs)

_transform_cpu(**kwargs)[source]: Respective immediate transform mocked function for local CPU(s).

_transform_gpu(**kwargs)[source]: Respective immediate transform mocked function for local GPU(s).

Parameters:: flatten (bool)

class dasf.feature_extraction.GetSubDataframe(percent)[source]

Get the first x% samples from the dataset.

Parameters

percentfloat: Percentage of the samples to get from the dataframe.

__percent

transform(X)[source]

Parameters:: percent (float)

class dasf.feature_extraction.SampleDataframe(percent)[source]

Bases: dasf.transforms.base.Transform

Return a subset with random samples of the original dataset.

Parameters

percentfloat: Percentage of the samples to get from the dataset.

__percent

transform(X)[source]

Returns a subset with random samples from the dataset X.

Parameters

XAny: The dataset.

Returns

Any: The sampled subset.

Parameters:: percent (float)