sample_odds_ratio#

rojak.turbulence.metrics.sample_odds_ratio(first_var: DataArray, second_var: DataArray, sum_over: str, *, use_log: bool = True) → DataArray[source]#

Sample odds ratio between two binary variables

An odds ratio (OR) is a statistic that quantifies the strength of the association between two events, A and B. The odds ratio is defined as the ratio of the odds of event A taking place in the presence of B, and the odds of A in the absence of B. Definition from Wikipedia.

Parameters:

  • first_var (DataArray) – A binary xarray DataArray representing the exposure variable

  • second_var (DataArray) – A binary xarray DataArray representing the outcome variable

  • sum_over (str) – Name of the dimension to sum over when computing the 2x2 contingency table. This dimension will be aggregated to obtain cell counts (n_00, n_01, n_10, n_11).

  • use_log (bool) – If True (default), returns the natural logarithm of the odds ratio. Set to False to return the raw odds ratio. Default is True.

Returns:

The sample odds ratio (or log odds ratio if use_log=True).

Dimensions are preserved except for the summed dimension.

Return type:

xr.DataArray

Notes

  • OR > 1 indicates increased odds of outcome with exposure

  • OR < 1 indicates decreased odds of outcome with exposure

  • OR = 1 indicates no association between exposure and outcome

  • The log odds ratio is often preferred for statistical analysis as it has better distributional properties