statsmodels.stats.contingency_tables.Table

class statsmodels.stats.contingency_tables.Table(table, shift_zeros=True)[source]

A two-way contingency table.

Parameters
  • table (array-like) – A contingency table.

  • shift_zeros (boolean) – If True and any cell count is zero, add 0.5 to all values in the table.

table_orig

The original table is cached as table_orig.

Type

array-like

marginal_probabilities[source]

The estimated row and column marginal distributions.

Type

tuple of two ndarrays

independence_probabilities[source]

Estimated cell probabilities under row/column independence.

Type

ndarray

fittedvalues[source]

Fitted values under independence.

Type

ndarray

resid_pearson[source]

The Pearson residuals under row/column independence.

Type

ndarray

standardized_resids[source]

Residuals for the independent row/column model with approximate unit variance.

Type

ndarray

chi2_contribs[source]

The contribution of each cell to the chi^2 statistic.

Type

ndarray

local_logodds_ratios

The local log odds ratios are calculated for each 2x2 subtable formed from adjacent rows and columns.

Type

ndarray

local_oddsratios[source]

The local odds ratios are calculated from each 2x2 subtable formed from adjacent rows and columns.

Type

ndarray

cumulative_log_oddsratios[source]

The cumulative log odds ratio at a given pair of thresholds is calculated by reducing the table to a 2x2 table based on dichotomizing the rows and columns at the given thresholds. The table of cumulative log odds ratios presents all possible cumulative log odds ratios that can be formed from a given table.

Type

ndarray

cumulative_oddsratios[source]

The cumulative odds ratios are calculated by reducing the table to a 2x2 table based on cutting the rows and columns at a given point. The table of cumulative odds ratios presents all possible cumulative odds ratios that can be formed from a given table.

Type

ndarray

See also

statsmodels.graphics.mosaicplot.mosaic, scipy.stats.chi2_contingency

Notes

The inference procedures used here are all based on a sampling model in which the units are independent and identically distributed, with each unit being classified with respect to two categorical variables.

References

Definitions of residuals:

https://onlinecourses.science.psu.edu/stat504/node/86

Methods

chi2_contribs()

Returns the contributions to the chi^2 statistic for independence.

cumulative_log_oddsratios()

Returns cumulative log odds ratios.

cumulative_oddsratios()

Returns the cumulative odds ratios for a contingency table.

fittedvalues()

Returns fitted cell counts under independence.

from_data(data[, shift_zeros])

Construct a Table object from data.

independence_probabilities()

Returns fitted joint probabilities under independence.

local_log_oddsratios()

Returns local log odds ratios.

local_oddsratios()

Returns local odds ratios.

marginal_probabilities()

Estimate marginal probability distributions for the rows and columns.

resid_pearson()

Returns Pearson residuals.

standardized_resids()

Returns standardized residuals under independence.

test_nominal_association()

Assess independence for nominal factors.

test_ordinal_association([row_scores, …])

Assess independence between two ordinal variables.