pyCADD.Dance.algorithm package

Submodules

pyCADD.Dance.algorithm.DL module

pyCADD.Dance.algorithm.consensus module

class pyCADD.Dance.algorithm.consensus.Average(lower_is_better: bool = False)[source]

Bases: _Consensus

Arithmetic mean consensus model.

Calculates arithmetic mean of molecular descriptor values across conformations. Arithmetic mean is defined as: mean = sum(x_i) / count

__init__(lower_is_better: bool = False) → None[source]

Initialize arithmetic mean consensus algorithm.

Parameters:

lower_is_better (bool) – Whether lower values indicate better performance.

fit(X: DataFrame, y: Series = None, ignore_nan: bool = True) → None[source]

Calculate arithmetic mean across conformations.

Parameters:

  • X (DataFrame) – Feature data containing molecular descriptors across conformations.

  • y (Series, optional) – Data labels (not used in calculation, exists for compatibility).

  • ignore_nan (bool) – Whether to ignore missing values (all 0 values are treated as missing).

class pyCADD.Dance.algorithm.consensus.Mean(lower_is_better: bool = False)[source]

Bases: Average

Arithmetic mean consensus model.

Alias for the Average class providing the same arithmetic mean functionality.

class pyCADD.Dance.algorithm.consensus.Geo_Average(lower_is_better: bool = False)[source]

Bases: _Consensus

Geometric mean consensus model.

Calculates geometric mean of molecular descriptor values across conformations. Geometric mean is defined as: geo_mean = (product(x_i))^(1/n)

__init__(lower_is_better: bool = False) → None[source]

Initialize geometric mean consensus algorithm.

Parameters:

lower_is_better (bool) – Whether lower values indicate better performance.

fit(X: DataFrame, y: Series = None, ignore_nan: bool = True) → None[source]

Calculate geometric mean across conformations.

Takes absolute values for calculation, then applies sign.

Parameters:

  • X (DataFrame) – Feature data containing molecular descriptors across conformations.

  • y (Series, optional) – Data labels (not used in calculation, exists for compatibility).

  • ignore_nan (bool) – Whether to ignore missing values (all 0 values are treated as missing).

class pyCADD.Dance.algorithm.consensus.GeoMean(lower_is_better: bool = False)[source]

Bases: Geo_Average

Geometric mean consensus model.

Alias for the Geo_Average class providing the same geometric mean functionality.

class pyCADD.Dance.algorithm.consensus.Minimum(lower_is_better: bool = False)[source]

Bases: _Consensus

Minimum value consensus model.

Selects the minimum value across conformations for each molecule.

__init__(lower_is_better: bool = False) → None[source]

Initialize minimum value consensus algorithm.

Parameters:

lower_is_better (bool) – Whether lower values indicate better performance.

fit(X: DataFrame, y: Series = None, ignore_nan: bool = True) → None[source]

Calculate minimum values across conformations.

Parameters:

  • X (DataFrame) – Feature data containing molecular descriptors across conformations.

  • y (Series, optional) – Data labels (not used in calculation, exists for compatibility).

  • ignore_nan (bool) – Whether to ignore missing values (all 0 values are treated as missing).

class pyCADD.Dance.algorithm.consensus.Maximum(lower_is_better: bool = False)[source]

Bases: _Consensus

Maximum value consensus model.

Selects the maximum value across conformations for each molecule.

__init__(lower_is_better: bool = False) → None[source]

Initialize maximum value consensus algorithm.

Parameters:

lower_is_better (bool) – Whether lower values indicate better performance.

fit(X: DataFrame, y: Series = None, ignore_nan: bool = True) → None[source]

Calculate maximum values across conformations.

Parameters:

  • X (DataFrame) – Feature data containing molecular descriptors across conformations.

  • y (Series, optional) – Data labels (not used in calculation, exists for compatibility).

  • ignore_nan (bool) – Whether to ignore missing values (all 0 values are treated as missing).

pyCADD.Dance.algorithm.consensus.average(data: DataFrame, method: Literal['ave', 'geo'] = 'ave') → Series[source]

Calculate average values across molecular conformations.

Provides functional interface for average calculations: - ‘ave’: Arithmetic mean - ‘geo’: Geometric mean

Parameters:

  • data (DataFrame) – DataFrame containing molecular data to calculate averages from.

  • method (str) – Average calculation method (‘ave’ for arithmetic mean | ‘geo’ for geometric mean).

Returns:

Series containing average values.

Raises:

RuntimeError – If method is not ‘ave’ or ‘geo’.

pyCADD.Dance.algorithm.consensus.minimum(data: DataFrame) → Series[source]

Extract minimum values (best scores) from DataFrame columns.

Useful for extracting optimal scores such as best docking scores.

Parameters:

data (DataFrame) – DataFrame containing molecular data to extract minimums from.

Returns:

Series containing minimum values for each row.

pyCADD.Dance.algorithm.consensus.maximum(data: DataFrame) → Series[source]

Extract maximum values from DataFrame columns.

Parameters:

data (DataFrame) – DataFrame containing molecular data to extract maximums from.

Returns:

Series containing maximum values for each row.

pyCADD.Dance.algorithm.consensus.std(data: DataFrame, axis: int = 1) → Series[source]

Calculate standard deviation.

Parameters:

  • data (DataFrame) – DataFrame containing molecular data to calculate standard deviation from.

  • axis (int) – Calculation axis (0 for rows, 1 for columns).

Returns:

Series containing standard deviation values.

pyCADD.Dance.algorithm.default_params module

Module contents