A Python Package for Grappling with Missing Data
summary
Auto impute is a pip‑installable Python package with a scikit‑learn style API that runs end‑to‑end missing‑data workflows: explore missingness, run visualizations of missing locations, try multiple imputation strategies (mean, least‑squares regression, predictive mean matching), and pool results into regression estimates. Examples use synthetic 500‑row datasets with ~40% missingness and 5 imputations per method; results show mean and regression imputation shrink variance and bias coefficients while PMM preserves distributions and between‑imputation variability, and PMM is the default for continuous variables.
AI-written summary · about
chapters (14)
- 0:00Auto impute package intro
- 3:28Missing data challenges
- 8:25Missing data mechanisms
- 13:50Imputation considerations
- 18:18Package features and setup
- 21:54Demo overview and data setup
- 22:15Visualizing missingness
- 24:44Mean imputation
- 27:36Regression imputation
- 29:01Predictive mean matching
- 32:09Imputing MAR data
- 38:11Choosing imputation method
- 39:18Resources and future work
- 40:01Q&A
related talks
26mThe industrial challenge of missing data
27mFrom Data to Insights in Minutes: Accelerating Predictive Modelling (in Python) with AutoML
32mpyGAM: balancing interpretability and predictive power using...
45mPython useRs
01h 53mAnalyze your data in Python
01h 24mInpsect and try to interpret your scikit-learn machine-learning models
29mPIP Internals
44mMachine Learning Made Easy With PyCaret