CSCAR workshop materials

Python notebooks by topic:

Linear regression

• Regression analysis of the US house price index (notebook)

• Regression analysis of the NHANES blood pressure data (notebook)

Multiple testing and False Discovery Rates

• Differential expression analysis of gene expression data (notebook)

• Simulation study of FDR methods (notebook)

• FDR and large scale inference (notebook

Predictive modeling

• Cross validation and ROC curves for predicting diabetes status (notebook)

Dotplots (forest plots)

• GDP by country (notebook, data)

• Gas prices (notebook)

Principal Component Analysis

• PCA of fertility rates by country (notebook)

Regularized model fitting

• L1 regularized logistic regression example with simulated data (notebook)

• L1 regularized logistic regression with data from medical utilization studies (notebook)

• L1 linear regression for a diabetes data set (notebook)

Generalized Linear Models

• Logistic regression power analysis using simulation (notebook)

• Relative risk logistic regression (notebook)

• Regression graphics (notebook)

• Gamma GLM (notebook)

• Logistic regression analysis of low birth weight data (notebook)

• Poisson and negative binomial regression analysis of fish abundance data (notebook)

Scatterplot smoothing (lowess)

• Cross sectional smoothing of NHANES data (notebook)

Survival analysis

• Survival analysis of NHANES 3 data (notebook)

• Basic proportional hazards regression in Statsmodels, R, and Stata (notebook)

• Diagnostics for proportional hazards regression models (notebook)

• Simulated data example for dependent events (notebook)

• Prediction in proportional hazards models (notebook)

Covariance matrix estimation

• Simulation study using "corr_nearest" (notebook)

Linear Mixed Models

• Two examples using both R (LME4) and Statsmodels (notebook)

• Regression analysis of healthcare spending in Vietnam, using mixed models and GEE. Stata results provided for comparison (notebook)

Generalized Estimating Equations (GEE)

• GEE analysis of longitudinal CD4 counts (notebook)

• GEE Poisson model for repeated measures of epileptic seizure counts (notebook)

• GEE Gaussian and Poisson models for repeated measures of disease incidence in herds of cattle (notebook, data set)

• GEE for discrete ordinal test score data with cluster sampling (notebook)

• GEE score test simulation study (notebook)

• GEE for repeated measures of outcomes that are proportions (notebook)

• GEE analyses of student test score data with nested dependence (notebook)

• GEE simulation study for data with nested dependence (notebook)

• GEE simulation study for Poisson regression with overdispersion (notebook)

• GEE simulation study for regression with nominal data (notebook)

• Baseline adjustment and GEE (notebook)

Missing data (MICE multiple imputation)

• Chain data (notebook)

• Survival analysis (notebook)

• Simulation study (notebook)

Data management with Pandas

• Data structures in Python

Graphics and data visualization with Matplotlib and Bokeh

• Data visualization overview

Pandas case studies (CMS data analyses):

• Subset the data by state (notebook)

• Drill-down: identify providers receiving high payments (notebook)

• Compare provider types by state (notebook)

• Associations beween provider type and service type (notebook)

Survival analysis with Statsmodels

• Survival analysis of NHANES 3 data (notebook)

Sklearn

• notebook