Workflow module

ConditionalDependence

Class containing methods for calculating conditional dependence.

__init__(y, z)

Initialize and Validate a ConditionalDependence object.

You can then pass any X values to compute_conditional_dependence and compute_conditional_dependence_1d

Parameters:

Name	Type	Description	Default
y	npt.ArrayLike	A single list or 1D array or a pandas Series.	required
z	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	required

Raises:

Type	Description
ValueError	If y is not 1d.
ValueError	If z is not 1d or 2d.
ValueError	If y and z have different lengths.
ValueError	If there are <= 2 valid y values.

compute_conditional_dependence(x=None)

Compute conditional dependence coefficient based on

Azadkia and Chatterjee (2021). "A simple measure of conditional dependence", Annals of Statistics

If X is passed, computes T(Y, Z|X) where T is the conditional dependence coefficient. Otherwise, computes T(Y, Z).

Conditional Dependence Coefficient lies between 0 and 1, and is

0 if Y is completely independent of Z|X
1 if Y is a measurable function of Z|X

Parameters:

Name	Type	Description	Default
x	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	None

Returns:

Name	Type	Description
float		Conditional Dependence Coefficient.

Raises:

Type	Description
ValueError	If x is passed, and not same number of rows as y.

compute_conditional_dependence_1d(x=None)

Computes conditional dependence of y on each column of z individually.

Use when you want to compute T(Y, Z_j|X) for each column of Z.

Parameters:

Name	Type	Description	Default
x	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	None

Returns:

Name	Type	Description
dict		Keys are column names (or indices if x is not a pandas object), and values are conditional dependence coefficients.

Raises:

Type	Description
ValueError	If x is passed, and does not have same number of rows as y.

compute_conditional_dependence(y, z, x=None)

Compute conditional dependence coefficient based on

Azadkia and Chatterjee (2021). "A simple measure of conditional dependence", Annals of Statistics

If X is passed, computes T(Y, Z|X) where T is the conditional dependence coefficient. Otherwise, computes T(Y, Z).

Conditional Dependence Coefficient lies between 0 and 1, and is

0 if Y is completely independent of Z|X
1 if Y is a measurable function of Z|X

Parameters:

Name	Type	Description	Default
y	npt.ArrayLike	A single list or 1D array or a pandas Series.	required
z	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	required
x	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	None

Returns:

Name	Type	Description
float	float	Conditional Dependence Coefficient.

Raises:

Type	Description
ValueError	If y is not 1d.
ValueError	If z is not 1d or 2d.
ValueError	If y and z have different lengths.
ValueError	If there are <= 2 valid y values.
ValueError	If x is passed, and not same number of rows as y.

compute_conditional_dependence_1d(y, z, x=None)

Computes conditional dependence of y on each column of z individually.

Use when you want to compute T(Y, Z_j|X) for each column of Z.

Parameters:

Name	Type	Description	Default
y	npt.ArrayLike	A single list or 1D array or a pandas Series.	required
z	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	required
x	npt.ArrayLike	A single list or list of lists or 1D/2D numpy array or pd.Series or pd.DataFrame.	None

Returns:

Name	Type	Description
dict	Dict[Union[str, int], float]	Keys are column names (or indices if x is not a pandas object), and values are conditional dependence coefficients.

Raises:

Type	Description
ValueError	If y is not 1d.
ValueError	If z is not 1d or 2d.
ValueError	If y and z have different lengths.
ValueError	If there are <= 2 valid y values.
ValueError	If x is passed, and does not have the same number of rows as y.