API Reference¶

Dichotomous variables¶

correlation coefficients for dichotomous (binary) variables

:py:obj:`ordinalcorr.biserial <ordinalcorr.biserial>`\ (x, y)	Compute the biserial correlation coefficient between a continuous variable :math:`x` and a dichotomized variable :math:`y\in\{0,1\}`, assuming :math:`y` is generated by discretizing a latent continuous variable.
:py:obj:`ordinalcorr.point_biserial <ordinalcorr.point_biserial>`\ (x, y)	Compute the point-biserial correlation between a continuous variable :math:`x` and a dichotomous variable :math:`y\in\{0,1\}`, assuming :math:`y` is a true dichotomous variable (i.e. :math:`y` is not generated from a latent continuous variable).
:py:obj:`ordinalcorr.tetrachoric <ordinalcorr.tetrachoric>`\ (x, y)	Estimate the tetrachoric correlation coefficient between two dichotomous variables.

correlation coefficients for polytomous (ordered categorical) variables

:py:obj:`ordinalcorr.polychoric <ordinalcorr.polychoric>`\ (x, y)	Estimate the polychoric correlation coefficient between two ordinal variables.
:py:obj:`ordinalcorr.polyserial <ordinalcorr.polyserial>`\ (x, y)	Estimate the polyserial correlation coefficient between a continuous variable :math:`x` and an ordinal variable :math:`y` using the two-step maximum likelihood estimation.

:py:obj:ordinalcorr.hetcor <ordinalcorr.hetcor>\ (data[, n_categories, ...])

Estimate the heterogeneous correlation matrix.