nimare.correct.FDRCorrector

class FDRCorrector(method='indep', alpha=0.05, **kwargs)

Bases: Corrector

Perform false discovery rate correction on a meta-analysis.

Parameters

• method (str, optional) – The FDR correction to use. Either ‘indep’ (for independent or positively correlated values) or ‘negcorr’ (for general or negatively correlated tests). Default is ‘indep’.

• alpha (float, optional) – The FDR correction rate to use. Default is 0.05.

Notes

This corrector supports a small number of internal FDR correction methods, but can also use special methods implemented within individual Estimators. To determine what methods are available for the Estimator you’re using, use inspect(). Estimators have special methods following the naming convention correct_[correction-type]_[method] (e.g., correct_fdr_indep).

Methods

correct_fdr_indep(p)	Perform Benjamini-Hochberg FDR correction.
correct_fdr_negcorr(p)	Perform Benjamini-Yekutieli FDR correction.
inspect(result)	Identify valid 'method' values for a MetaResult object.
transform(result)	Apply the multiple comparisons correction method to a MetaResult object.

correct_fdr_indep(p)

Perform Benjamini-Hochberg FDR correction.

This correction is based on the one described in Benjamini and Hochberg1. This method is not universally appropriate. It works well for tests that are independent, or which are positively correlated.

Warning

Do not call this method directly. Call transform() with method='indep' instead.

New in version 0.0.12.

Parameters

p (numpy.ndarray) – A 1D array of p values.

Returns

A 1D array of adjusted p values.

Return type

numpy.ndarray

References

1

Yoav Benjamini and Yosef Hochberg. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological), 57(1):289–300, 1995. URL: https://doi.org/10.1111/j.2517-6161.1995.tb02031.x, doi:10.1111/j.2517-6161.1995.tb02031.x.

See also

pymare.stats.fdr

correct_fdr_negcorr(p)

Perform Benjamini-Yekutieli FDR correction.

This correction is based on the one described in Benjamini and Yekutieli2. It is most appropriate for tests that are negatively correlated.

Warning

Do not call this method directly. Call transform() with method='negcorr' instead.

New in version 0.0.12.

Parameters

p (numpy.ndarray) – A 1D array of p values.

Returns

A 1D array of adjusted p values.

Return type

numpy.ndarray

Notes

The difference between the Benjamini-Yekutieli and Benjamini-Hochberg methods is that Benjamini-Yekutieli includes an additional term, c(m). When the tests are independent or positively correlated, c(m) is 1 (and thus has no effect). In cases of other forms of dependence, c(m) has an effect.

References

2

Yoav Benjamini and Daniel Yekutieli. The control of the false discovery rate in multiple testing under dependency. The Annals of Statistics, 29(4):1165 – 1188, 2001. URL: https://doi.org/10.1214/aos/1013699998, doi:10.1214/aos/1013699998.

See also

pymare.stats.fdr

classmethod inspect(result)

Identify valid ‘method’ values for a MetaResult object.

In addition to returning a list of valid values, this method will also print out those values, divided by the value type (Estimator or generic).

Parameters

result (MetaResult) – Object for which valid correction methods (i.e., ‘method’ values) will be identified.

Returns

List of valid ‘method’ values for the Corrector+Estimator combination, including both non-specific methods and Estimator-specific ones.

Return type

list

transform(result)

Apply the multiple comparisons correction method to a MetaResult object.

Parameters

result (MetaResult) – MetaResult generated by an Estimator to be corrected for multiple comparisons.

Returns

result – MetaResult with new corrected maps added.

Return type

MetaResult

Examples using nimare.correct.FDRCorrector

The Corrector class

Simulate data for coordinate based meta-analysis