pymargins.ImputationDiagnostic¶
- class pymargins.ImputationDiagnostic(n_imputations: int, fmi: ndarray | float, relative_efficiency: ndarray | float, df: ndarray | float, within_var: ndarray | float, between_var: ndarray | float, total_var: ndarray | float, riv: ndarray | float)¶
Diagnostic information from Rubin pooling.
Each per-component field is a Python
floatfor a scalar estimand and annp.ndarray(shaped likeMarginsResult.estimate) for a vector one.- n_imputations¶
Number of imputations pooled.
- Type:
int
- fmi¶
Fraction of missing information per component.
- Type:
np.ndarray | float
- relative_efficiency¶
Relative efficiency per component.
- Type:
np.ndarray | float
- df¶
Degrees of freedom per component.
- Type:
np.ndarray | float
- within_var¶
Within-imputation variance per component.
- Type:
np.ndarray | float
- between_var¶
Between-imputation variance per component.
- Type:
np.ndarray | float
- total_var¶
Total variance per component.
- Type:
np.ndarray | float
- riv¶
Relative increase in variance per component.
- Type:
np.ndarray | float
- __init__(n_imputations: int, fmi: ndarray | float, relative_efficiency: ndarray | float, df: ndarray | float, within_var: ndarray | float, between_var: ndarray | float, total_var: ndarray | float, riv: ndarray | float) None¶
Methods
__init__(n_imputations, fmi, ...)footer()Return a one-line diagnostic summary for summary() footers.
Attributes