pylimma.top_table_f

pylimma.top_table_f(fit, number=10, genelist=None, adjust_method='BH', sort_by='F', p_value=1.0, fc=None, lfc=None, *, coef_idx=None, resort_by=None, _genelist_explicit=None)[source]

Top table for multiple coefficients ranked by F-statistic.

Port of R’s limma::topTableF. Positional arguments follow R’s order exactly; coef_idx and resort_by are pylimma-only extensions and keyword-only.

Parameters:

  • fit (dict) – Fit object from e_bayes() containing F-statistics.

  • number (int, default 10) – Maximum number of genes to return.

  • genelist (DataFrame, list, or array, optional) –

    Gene annotations to include in output. When None, uses fit.get("genes") (matches R’s genelist=fit$genes default).

    For AnnData input, fit[“genes”] only carries var_names; pass genelist=adata.var explicitly to merge annotation columns (e.g. symbol, chromosome) into the output table.

  • adjust_method (str, default "BH") – Multiple-testing adjustment method.

  • sort_by (str, default "F") – Column to sort by. "F" or "none" (R).

  • p_value (float, default 1.0) – Adjusted p-value cutoff.

  • fc (float, optional) – Fold-change cutoff. If given, sets lfc = log2(fc). Must be >= 1.

  • lfc (float, optional) – Log fold-change cutoff. Defaults to 0 when both fc and lfc are None (R’s NULL).

  • coef_idx (list of int, optional (keyword-only)) – Indices of coefficients to include. Defaults to all columns of fit["coefficients"], matching R’s topTableF which always uses the whole coefficient matrix.

  • resort_by (str, optional (keyword-only)) – pylimma extension: secondary sort column applied after sort_by + truncation.

  • _genelist_explicit (bool | None)

Returns:

Table of top genes ranked by F-statistic.

Return type:

DataFrame