Plot continuous outcome and its related variables in SAS
Bhagirathi Dash
10/16/2019
- Dataset and its variables
- Import the dataset
- Describe and display its contents
- Summary statistics for continuous variables
- One-way frequency tables for categorical variables
- PLOT CONTINUOUS OUTCOME AND ITS RELATED VARIABLES
- Plot a continuous variable (without a grouping variable)
- Plot a continuous variable (with a grouping variable)
- Plot the mean of a continuous variable (with a grouping a variable)
- Scatter plots
- Bivariate scatter plots ( without a grouping variable)
- Bivariate scatter plots (with a grouping variable)
- Bivariate scatter plots (with a grouping variable) using proc sgpanel
- Bivariate scatter plots (with two grouping variables) using proc sgpanel
- Independent multiple scatter plots (without a grouping variable) using proc sgscatter
- Independent multiple scatter plots (with a grouping variable) using proc sgscatter
- Comparative panels of independent multiple scatter plots (with a grouping variable) using proc sgscatter
- Matrix of scatter plots (with a grouping variable) using proc sgscatter
Dataset and its variables
We will use the birthwt dataset from the MASS library in R.
| # | variable name | variable label | coded levels |
|---|---|---|---|
| 1 | low | indicator of birth weight less than 2.5 kg | 0, 1 |
| 2 | age | mother’s age in years | continuous variable |
| 3 | lwt | mother’s weight in pounds at last menstrual period | continuous variable |
| 4 | race | mother’s race (1 = white, 2 = black, 3 = other) | 1, 2, 3 |
| 5 | smoke | smoking status during pregnancy | 0, 1 |
| 6 | ptl | number of previous premature labours | 0, 1, 2, 3 |
| 7 | ht | history of hypertension | 0, 1 |
| 8 | ui | presence of uterine irritability | 0, 1 |
| 9 | ftv | number of physician visits during the first trimester | 0, 1, 2, 3, 4, 6 |
| 10 | bwt | birth weight in grams | continuous variable |
Import the dataset
%let path=/folders/myfolders/birthwt;SAS Connection established. Subprocess id is 2929<!DOCTYPE html PUBLIC “-//W3C//DTD HTML 4.01//EN” “http://www.w3.org/TR/html4/strict.dtd”>
34 ods listing close;ods html5 (id=saspy_internal) file=stdout options(bitmap_mode='inline') device=svg style=HTMLBlue; ods
34 ! graphics on / outputfmt=png;
NOTE: Writing HTML5(SASPY_INTERNAL) Body file: STDOUT
35
36 %let path=/folders/myfolders/birthwt;
37
38 ods html5 (id=saspy_internal) close;ods listing;
39
proc import datafile="&path/birthwt.csv" out=ibwt dbms=csv replace;<!DOCTYPE html PUBLIC “-//W3C//DTD HTML 4.01//EN” “http://www.w3.org/TR/html4/strict.dtd”>
49 ods listing close;ods html5 (id=saspy_internal) file=stdout options(bitmap_mode='inline') device=svg style=HTMLBlue; ods
49 ! graphics on / outputfmt=png;
NOTE: Writing HTML5(SASPY_INTERNAL) Body file: STDOUT
50
51 proc import datafile="&path/birthwt.csv" out=ibwt dbms=csv replace;
52
53 ods html5 (id=saspy_internal) close;ods listing;
54
Describe and display its contents
proc contents data=ibwt varnum;
run;The CONTENTS Procedure
| Data Set Name | WORK.IBWT | Observations | 189 |
|---|---|---|---|
| Member Type | DATA | Variables | 10 |
| Engine | V9 | Indexes | 0 |
| Created | 10/14/2019 21:04:41 | Observation Length | 80 |
| Last Modified | 10/14/2019 21:04:41 | Deleted Observations | 0 |
| Protection | Compressed | NO | |
| Data Set Type | Sorted | NO | |
| Label | |||
| Data Representation | SOLARIS_X86_64, LINUX_X86_64, ALPHA_TRU64, LINUX_IA64 | ||
| Encoding | utf-8 Unicode (UTF-8) |
| Engine/Host Dependent Information | |
|---|---|
| Data Set Page Size | 65536 |
| Number of Data Set Pages | 1 |
| First Data Page | 1 |
| Max Obs per Page | 817 |
| Obs in First Data Page | 189 |
| Number of Data Set Repairs | 0 |
| Filename | /tmp/SAS_work2F0900000E52_localhost.localdomain/ibwt.sas7bdat |
| Release Created | 9.0401M6 |
| Host Created | Linux |
| Inode Number | 671803 |
| Access Permission | rw-r–r– |
| Owner Name | sasdemo |
| File Size | 128KB |
| File Size (bytes) | 131072 |
| Variables in Creation Order | |||||
|---|---|---|---|---|---|
| # | Variable | Type | Len | Format | Informat |
| 1 | low | Num | 8 | BEST12. | BEST32. |
| 2 | age | Num | 8 | BEST12. | BEST32. |
| 3 | lwt | Num | 8 | BEST12. | BEST32. |
| 4 | race | Num | 8 | BEST12. | BEST32. |
| 5 | smoke | Num | 8 | BEST12. | BEST32. |
| 6 | ptl | Num | 8 | BEST12. | BEST32. |
| 7 | ht | Num | 8 | BEST12. | BEST32. |
| 8 | ui | Num | 8 | BEST12. | BEST32. |
| 9 | ftv | Num | 8 | BEST12. | BEST32. |
| 10 | bwt | Num | 8 | BEST12. | BEST32. |
proc print data=ibwt (obs=5);
run;| Obs | low | age | lwt | race | smoke | ptl | ht | ui | ftv | bwt |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0 | 19 | 182 | 2 | 0 | 0 | 0 | 1 | 0 | 2523 |
| 2 | 0 | 33 | 155 | 3 | 0 | 0 | 0 | 0 | 3 | 2551 |
| 3 | 0 | 20 | 105 | 1 | 1 | 0 | 0 | 0 | 1 | 2557 |
| 4 | 0 | 21 | 108 | 1 | 1 | 0 | 0 | 1 | 2 | 2594 |
| 5 | 0 | 18 | 107 | 1 | 1 | 0 | 0 | 1 | 0 | 2600 |
Summary statistics for continuous variables
proc means data=ibwt maxdec=2;
var age lwt bwt;
run;The MEANS Procedure
| Variable | N | Mean | Std Dev | Minimum | Maximum |
|---|---|---|---|---|---|
|
age
lwt
bwt
|
189
189
189
|
23.24
129.81
2944.59
|
5.30
30.58
729.21
|
14.00
80.00
709.00
|
45.00
250.00
4990.00
|
One-way frequency tables for categorical variables
proc freq data=ibwt;
tables low race smoke ptl ht ui ftv/nocumfreq nocumpercent nopercent;
run;The FREQ Procedure
| low | Frequency |
|---|---|
| 0 | 130 |
| 1 | 59 |
| race | Frequency |
|---|---|
| 1 | 96 |
| 2 | 26 |
| 3 | 67 |
| smoke | Frequency |
|---|---|
| 0 | 115 |
| 1 | 74 |
| ptl | Frequency |
|---|---|
| 0 | 159 |
| 1 | 24 |
| 2 | 5 |
| 3 | 1 |
| ht | Frequency |
|---|---|
| 0 | 177 |
| 1 | 12 |
| ui | Frequency |
|---|---|
| 0 | 161 |
| 1 | 28 |
| ftv | Frequency |
|---|---|
| 0 | 100 |
| 1 | 47 |
| 2 | 30 |
| 3 | 7 |
| 4 | 4 |
| 6 | 1 |
PLOT CONTINUOUS OUTCOME AND ITS RELATED VARIABLES
Lets's assume that we are intersested to see how 'bwt' could be pedicted on other variables (except 'low').Plot a continuous variable (without a grouping variable)
Histogram, kernel density curve, normal density curve and box plot
ods noproctitle;
ods layout Start rows=1 columns=3;
ods region row=1 column=1;
proc sgplot data=ibwt;
histogram bwt / scale=count;
density bwt / type=normal;
density bwt / type=kernel;
title "Histogram for bwt";
run;
title;
ods region row=1 column=2;
proc univariate data=ibwt;
ods select qqplot;
var bwt;
qqplot bwt / normal(mu=est sigma=est);
*title "Qq plot for bwt";
run;
title;
ods region row=1 column=3;
proc sgplot data=ibwt;
hbox bwt;
title "Box plot for bwt";
run;
title;
ods layout end;|
|
|
|
Plot a continuous variable (with a grouping variable)
Histogram, kernel density curve, normal density curve and box plot
ods layout Start rows=3 columns=2;
ods region row=1 column=1;
proc sgplot data=ibwt;
histogram bwt / group=race scale=count;
density bwt /group=race type=normal;
density bwt /group=race type=kernel;
title "Histograms for bwt by race (sgplot)";
run;
title;
ods region row=1 column=2;
proc sgpanel data=ibwt;
panelby race/ columns=3;
histogram bwt/scale=count;
density bwt/type=kernel;
density bwt/type=normal;
title "Histograms for bwt by race (sgpanel)";
run;
title;
ods region row=2 column=1;
proc sgplot data=ibwt;
hbox bwt/group=race;
title "Box plots for bwt by race (sgplot)";
run;
title;
ods region row=2 column=2;
proc sgpanel data=ibwt;
panelby race/ columns=3;
hbox bwt;
title "Box plots for bwt by race (sgpanel)";
run;
title;
ods region row=3 column=1;
proc sgplot data=ibwt;
hbox bwt/category=race;
title "Box plots for bwt by race (sgplot/category)";
run;
title;
ods layout end;|
|
|
|
|
|
|
|
proc sort data=ibwt;
by race;
run;
ods noproctitle;
proc univariate data=ibwt;
ods select qqplot;
by race;
var bwt;
qqplot bwt / normal(mu=est sigma=est);
*title "Qq plots for bwt by race";
run;
title;
Plot the mean of a continuous variable (with a grouping a variable)
ods graphics / noborder;
ods layout Start rows=3 columns=2;
ods region row=1 column=1;
proc sgplot data=ibwt;
dot race / response=bwt stat=mean limitstat=stddev numstd=1;
title "Dot plot for mean bwt by race";
run;
title;
ods region row=2 column=1;
proc sgplot data=ibwt;
vline race / response=bwt stat=mean;
title "Line plot for mean bwt by race";
run;
title;
ods region row=2 column=2;
proc sgplot data=ibwt;
vbar race / response=bwt stat=mean;
title "Bar plot for mean bwt by race";
run;
title;
ods region row=3 column=1;
proc sgpanel data=ibwt;
panelby smoke;
vbar race / response=bwt stat=mean;
title "Bar plot for mean bwt by race for various smoking groups";
run;
title;
ods region row=3 column=2;
proc sgpanel data=ibwt;
panelby smoke;
vbar race / response=bwt group=ftv groupdisplay=cluster stat=mean;
title "Bar plot for mean bwt grouped by race and ftv for various smoking groups";
run;
title;|
|
|
|
|
|
|
|
|
Scatter plots
Bivariate scatter plots ( without a grouping variable)
ods graphics / noborder;
ods layout Start rows=3 columns=2;
*scatter plot;
ods region row=1 column=1;
proc sgplot data=ibwt;
scatter x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled);
xaxis label="lwt";
yaxis label="bwt";
title "Jitter/scatter plot for bwt and lwt";
run;
title;
*scatter plot with ellipse;
ods region row=1 column=2;
proc sgplot data=ibwt;
scatter x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled);
ellipse x=lwt y=bwt;
xaxis label="lwt";
yaxis label="bwt";
title "Elipse fit for bwt and lwt";
run;
title;
*scatter plot with lowess line;
ods region row=2 column=1;
proc sgplot data=ibwt;
loess x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1);
xaxis label="lwt";
yaxis label="bwt";
title "Loess fit for bwt and lwt";
run;
title;
*fitted penalized B-spline curve;
ods region row=2 column=2;
proc sgplot data=ibwt;
pbspline x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1);
xaxis label="lwt";
yaxis label="bwt";
title "B-spline fit for bwt and lwt";
run;
title;
*scatter plot with linear regression line;
ods region row=3 column=1;
proc sgplot data=ibwt;
reg x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1);
xaxis label="lwt";
yaxis label="bwt";
title "Linear regression fit for bwt and lwt";
run;
title;
*scatter plot with quadratic regression line;
ods region row=3 column=2;
proc sgplot data=ibwt;
reg x=lwt y=bwt / markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1) degree=2;
xaxis label="lwt";
yaxis label="bwt";
title "Quadratic regression fit for bwt and lwt";
run;
title;
ods layout end;|
|
|
|
|
|
|
|
|
Bivariate scatter plots (with a grouping variable)
ods graphics / noborder;
ods layout Start rows=3 columns=2;
*scatter plot;
ods region row=1 column=1;
proc sgplot data=ibwt;
scatter x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled);
xaxis label="lwt";
yaxis label="bwt";
title "Jitter/scatter plot for bwt and lwt by race";
run;
title;
*scatter plot with ellipse;
ods region row=1 column=2;
proc sgplot data=ibwt;
scatter x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled);
ellipse x=lwt y=bwt/group=race;
xaxis label="lwt";
yaxis label="bwt";
title "Elipse fit for bwt and lwt by race";
run;
title;
*scatter plot with lowess line;
ods region row=2 column=1;
proc sgplot data=ibwt;
loess x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled);
xaxis label="lwt";
yaxis label="bwt";
title "Loess fit for bwt and lwt by race";
run;
title;
*fitted penalized B-spline curve;
ods region row=2 column=2;
proc sgplot data=ibwt;
pbspline x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled);
xaxis label="lwt";
yaxis label="bwt";
title "B-spline fit for bwt and lwt by race";
run;
title;
*scatter plot with linear regression line;
ods region row=3 column=1;
proc sgplot data=ibwt;
reg x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled);
xaxis label="lwt";
yaxis label="bwt";
title "Linear regression fit for bwt and lwt by race";
run;
title;
*scatter plot with quadratic regression line;
ods region row=3 column=2;
proc sgplot data=ibwt;
reg x=lwt y=bwt /group=race markerattrs=(symbol=circlefilled) degree=2;
xaxis label="lwt";
yaxis label="bwt";
title "Quadratic regression fit for bwt and lwt by race";
run;
title;
ods layout end;
|
|
|
|
|
|
|
|
|
Bivariate scatter plots (with a grouping variable) using proc sgpanel
ods graphics / noborder;
ods layout Start rows=2 columns=2;
ods region row=1 column=1;
proc sgpanel data=ibwt;
panelby race/ columns=3;
scatter x=lwt y=bwt/ markerattrs=(symbol=circlefilled) ;
ellispse x=lwt y=bwt;
title "Ellipse fit for bwt and lwt by race";
run;
title;
ods region row=1 column=2;
proc sgpanel data=ibwt;
panelby race/ columns=3;
loess x=lwt y=bwt/ markerattrs=(symbol=circlefilled) lineattrs=(color=red thickness=1) ;
title "Loess fit for bwt and lwt by race";
run;
title;
ods region row=2 column=1;
proc sgpanel data=ibwt;
panelby race/ columns=3;
pbspline x=lwt y=bwt/ markerattrs=(symbol=circlefilled) lineattrs=(color=red thickness=1) ;
title "B-spline fit for bwt and lwt by race";
run;
title;
ods region row=2 column=2;
proc sgpanel data=ibwt;
panelby race/ columns=3;
reg x=lwt y=bwt/ markerattrs=(symbol=circlefilled) lineattrs=(color=red thickness=1) ;
title "Linear regression fit for bwt and lwt by race";
run;
title;|
|
|
|
|
|
Bivariate scatter plots (with two grouping variables) using proc sgpanel
ods graphics / noborder;
ods layout Start rows=1 columns=2;
ods region row=1 column=1;
proc sgpanle data=ibwt;
panelby race smoke/ rows=3 columns=2;
reg x=lwt y=bwt/markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1);
title "Linear regression fit for lwt and bwt by race and smoke";
run;
title;
ods region row=1 column=2;
proc sgpanle data=ibwt;
panelby smoke race/ rows=3 columns=2;
reg x=lwt y=bwt/markerattrs=(color=blue symbol=circlefilled) lineattrs=(color=red thickness=1);
title "Linear regression fit for lwt and bwt by smoke and race";
run;
title;
ods layout end;|
|
|
Independent multiple scatter plots (without a grouping variable) using proc sgscatter
ods graphics / noborder;
ods layout Start rows=2 columns=2;
ods region row=1 column=1;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/ ellipse;
title "Scatter plot with elipse";
run;
title;
ods region row=1 column=2;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/ loess;
title "Scatter plot with loess fit";
run;
title;
ods region row=2 column=1;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/ pbspline;
title "Scatter plot with b-spline fit";
run;
title;
ods region row=2 column=2;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/ reg ellipse=(type=mean);
title "Scatter plot with linear regression fit and an elipse around the mean";
run;
title;
ods layout end;|
|
|
|
|
|
Independent multiple scatter plots (with a grouping variable) using proc sgscatter
ods graphics / noborder;
ods layout Start rows=2 columns=2;
ods region row=1 column=1;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/group=race ellipse;
title "Scatter plot with ellipse fit";
run;
title;
ods region row=1 column=2;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/group=race loess;
title "Scatter plot with loess fit";
run;
title;
ods region row=2 column=1;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/group=race pbspline;
title "Scatter plot with b-spline fit";
run;
title;
ods region row=2 column=2;
proc sgscatter data=ibwt;
plot bwt *(lwt age)/group=race reg;
title "Scatter plot with linear regression fit";
run;
title;
ods layout end;|
|
|
|
|
|
Comparative panels of independent multiple scatter plots (with a grouping variable) using proc sgscatter
ods graphics / noborder;
ods layout Start rows=2 columns=2;
ods region row=1 column=1;
proc sgscatter data=ibwt;
compare x=(age lwt) y=bwt/group=race ellipse;
title "Scatter plot with ellipse fit";
run;
title;
ods region row=1 column=2;
proc sgscatter data=ibwt;
compare x=(age lwt) y=bwt/group=race loess;
title "Scatter plot with linear loess fit";
run;
title;
ods region row=2 column=1;
proc sgscatter data=ibwt;
compare x=(age lwt) y=bwt/group=race pbspline;
title "Scatter plot with b-spline fit";
run;
title;
ods region row=2 column=2;
proc sgscatter data=ibwt;
compare x=(age lwt) y=bwt/group=race reg;
title "Scatter plot with linear regression fit";
run;
title;
ods layout end;|
|
|
|
|
|
Matrix of scatter plots (with a grouping variable) using proc sgscatter
proc sgscatter data=ibwt;
matrix age lwt bwt/group=race;
title "Matrix of scatter plots with a grouping variable";
run;
title;