Example 15 of documentation for PROC MI

Niveau de difficulté

Débutant

Publié le : 25/09/2022

The script begins with the generation of a synthetic dataset named 'Mono1' containing missing values in the 'y1' variable. Then, it uses PROC MI to impute these missing values. The chosen imputation method is monotone regression with displayed details. Particular consideration is given to Missing Not At Random (MNAR) data on 'y1', modeled according to the treatment group 'Trt' when 'Trt' is equal to '0'. The imputed dataset is saved in 'outex15'. PROC PRINT procedures are used to display an overview of the original and imputed data.

Analyse des données

Type : CREATION_INTERNE

The initial dataset 'Mono1' is created directly in the script via a DATA STEP using random number generation functions (rannor, ranuni). The 'outex15' dataset is the result of the imputation performed by PROC MI.

1 Bloc de code

DATA STEP Data

Explication :
This DATA STEP block creates the 'Mono1' dataset. It generates observations for two treatment groups (Trt=0 and Trt=1) with 'y0' and 'y1' variables. The 'y1' variable is conditionally made missing (value '.') for approximately 30% of observations, thus simulating a missing data scenario.

Copié !

1	DATA Mono1;
2	DO Trt=0 to 1;
3	DO j=1 to 5;
4	y0=10 + rannor(999);
5	y1= y0 + Trt + rannor(999);
6	IF (ranuni(999) < 0.3) THEN y1=.;
7	OUTPUT;
8	END; END;
9
10	DO Trt=0 to 1;
11	DO j=1 to 45;
12	y0=10 + rannor(999);
13	y1= y0 + Trt + rannor(999);
14	IF (ranuni(999) < 0.3) THEN y1=.;
15	OUTPUT;
16	END; END;
17	drop j;
18	RUN;

2 Bloc de code

PROC PRINT

Explication :
This PROC PRINT displays the first 10 observations of the 'Mono1' dataset to provide an overview of the data structure before imputation. Only the 'Trt', 'Y0', and 'Y1' variables are included in the output.

Copié !

1	PROC PRINT DATA=Mono1(obs=10);
2	var Trt Y0 Y1;
3	title 'First 10 Obs in the Trial Data';
4	RUN;

3 Bloc de code

PROC MI

Explication :
This PROC MI performs multiple imputation of missing values in the 'Mono1' dataset. It uses seed '14823' for reproducibility and generates 15 imputed datasets, stored in 'outex15'. The 'monotone reg' method is specified for monotone regression imputation, with the 'details' option for additional information. The 'mnar' clause indicates that 'y1' is Missing Not At Random, and its model is conditioned on 'Trt' being '0'.

Copié !

1	PROC MI DATA=Mono1 seed=14823 nimpute=15 out=outex15;
2	class Trt;
3	monotone reg (/details);
4	mnar model( y1 / modelobs= (Trt='0'));
5	var y0 y1;
6	RUN;

4 Bloc de code

PROC PRINT

Explication :
This PROC PRINT displays the first 10 observations of the imputed dataset 'outex15'. This allows visualizing the results of the multiple imputation performed by PROC MI.

Copié !

1
2	PROC PRINT
3	DATA=outex15(obs=10);
4	title 'First 10 Observations of the Imputed
5	Data Set';
6	RUN;
7

Ce matériel est fourni "tel quel" par We Are Cas. Il n'y a aucune garantie, expresse ou implicite, quant à la qualité marchande ou à l'adéquation à un usage particulier concernant le matériel ou le code contenu dans les présentes. We Are Cas n'est pas responsable des erreurs dans ce matériel tel qu'il existe maintenant ou existera, et We Are Cas ne fournit pas de support technique pour celui-ci.

Informations de Copyright : SAS SAMPLE LIBRARY

Documentation liée

Aucune documentation spécifique pour cette catégorie.