microdf documentation

microdf provides weighted data structures for survey microdata analysis in Python. MicroSeries and MicroDataFrame carry sampling weights inside the object, so the weights stay aligned with their rows through merges, filters, grouping and reindexing, and the weighted estimators — quantiles, variance, Gini, top shares, poverty rates — use documented conventions rather than ad hoc ones.

import microdf as mdf

df = mdf.MicroDataFrame({"income": [10_000, 30_000, 120_000]}, weights=[800, 1_200, 50])
df.income.median()  # 30000, weighted
df.income.gini()

Install with pip install microdf-python.

• Examples — worked analyses, and how weights survive a pipeline

• API reference — every weighted estimator and weight-handling method