BTable

BTable(
    df,
    vars,
    group,
    *,
    labels=None,
    digits=2,
    pdigits=3,
    vcov='iid',
    fixed_effects=None,
    stats=None,
    stats_labels=None,
    format_spec=None,
    hide_stats=False,
    counts_row_below=False,
    observed=False,
    notes='',
    **kwargs,
)

Balancing table: descriptive stats by group + per-variable p-values from group tests.

Inherits DTable to build the stats table, then adds a ‘p-value’ column: - For 2 groups: p-value of the single group indicator (t test). - For >2 groups: joint Wald test that all group indicators are zero. You can add fixed_effects and specify the vcov option, for instance to implement clustering (see pyfixest documentation).

Parameters

Name	Type	Description	Default
df	pd.DataFrame	Source data.	required
vars	list[str]	Variables to include.	required
group	str	Grouping column in df.	required
labels	dict	Variable labels (used for display and in notes).	None
digits	int	Rounding for stats in the DTable. Default 2.	2
pdigits	int	Rounding for p-values. Default 3.	3
vcov	str | dict	VCV for the p-value models (“iid”, “hetero”, “HC1”, “HC2”, “HC3” or {“CRV1”: “cluster”}). Default “iid”.	'iid'
fixed_effects	list[str] | None	Optional fixed effects for the p-value models (cosmetic in notes). Default None.	None
stats	list[str] | None	Stats for DTable (default [“mean”, “std”]).	None
stats_labels	dict[str, str] | None	Custom labels for stats (used in header/notes).	None
hide_stats	bool	Hide stats names in the header (list them in notes instead). Default False.	False
counts_row_below	bool	If True and balanced N, show a single counts row at bottom. Default False.	False
observed	bool	Only consider observed categories when grouping. Default False.	False
notes	str	Custom notes. If ““, a default is generated mentioning stats/SE/FE.	''
kwargs	dict	Passed through to DTable/MTable (e.g., caption, tab_label, rgroup_display).	{}