Fixing Missing Data in SAS: A Guide to Multiple Imputation with PROC MI and MCMC

Difficulty Level

Beginner

Published on : 27/03/2024

Expert Advice

Michael
Responsable de l'infrastructure Viya.

When using the MCMC method in PROC MI, the chain=multiple option creates a separate Markov chain for each imputation; for better convergence results in smaller datasets like Fitness1, it is often beneficial to check the displayinit and itprint output to ensure that the EM algorithm (used for the starting values) has converged properly before the imputation begins.

The script begins by creating a 'Fitness1' dataset containing physical fitness measurements (oxygen consumption, run time, pulse). Some of these measurements are missing. Then, the MI procedure is used to impute these missing values by generating 40 complete datasets. Imputation is based on a Markov Chain Monte Carlo (MCMC) method, with a specified seed to ensure reproducibility.

Data Analysis

Type : CREATION_INTERNE

The data is created directly within the script via a DATA step and the DATALINES statement. No external data source is required.

1 Code Block

DATA STEP Data

Explanation :
This block creates the 'Fitness1' table from data embedded in the code (datalines). The 'input' statement reads the Oxygen, RunTime, and RunPulse variables. The ' @@' specifier allows reading multiple observations from the same data line.

Copied!

1	DATA Fitness1;
2	INPUT Oxygen RunTime RunPulse @code_sas_json/8_SAS_Intro_ReadFile_MultiCol_@@.json;
3	DATALINES;
4	44.609 11.37 178 45.313 10.07 185
5	54.297 8.65 156 59.571 . .
6	49.874 9.22 . 44.811 11.63 176
7	. 11.95 176 . 10.85 .
8	39.442 13.08 174 60.055 8.63 170
9	50.541 . . 37.388 14.03 186
10	44.754 11.12 176 47.273 . .
11	51.855 10.33 166 49.156 8.95 180
12	40.836 10.95 168 46.672 10.00 .
13	46.774 10.25 . 50.388 10.08 168
14	39.407 12.63 174 46.080 11.17 156
15	45.441 9.63 164 . 8.92 .
16	45.118 11.08 . 39.203 12.88 168
17	45.790 10.47 186 50.545 9.93 148
18	48.673 9.40 186 47.920 11.50 170
19	47.467 10.50 170
20	;
21

2 Code Block

PROC MI

Explanation :
This block uses the Multiple Imputation (MI) procedure on the 'Fitness1' table. It is configured to generate 40 imputed datasets ('nimpute=40') using an MCMC method. The 'seed' option ensures reproducibility. The variables to be imputed are specified in the 'var' statement.