Lesson 8 - Topics

1
Lesson 8 - Topics

• Creating SAS datasets from procedures
• Using ODS and data steps to make reports
• Using PROC RANK
• Programs 14-15 in course notes
• LSB 41153

2
Making SAS Datasets From Procedures Output from
SAS PROCs can be put into SAS datasets 1. To do
further processing of the information from the
output 2. To reformat output to make a
report 3. To restructure original SAS dataset or
create new variables
3
Ways to Put Output into SAS Datasets

• Using OUTPUT statement available from many
  procedures
• Using ODS OUTPUT statement any output table
  can be put into a SAS dataset

4
Report We Want to Generate Quartiles of Weight
by Gender and Center sex clinic N
P25 P50 P75 Male A
9 180.0 190.0 208.0 Male
B 16 158.3 174.8 218.3
Male C 29 178.0 199.5
220.5 Male D 11 172.0
184.5 194.0 Female A 6
125.0 143.5 160.5 Female B
9 150.0 164.5 184.0 Female
C 6 132.5 134.3 138.5
Female D 6 131.0 137.5
148.5
5
Program 14 LIBNAME class C\SAS_Files' Will
use SAS dataset version of TOMHS data DATA wt
SET class.tomhsp (KEEPptid age sex clinic wtbl
wt12 ) wtchg wt12 -
wtbl RUN PROC FORMAT VALUE sexF 1 Male
2Female RUN
6
Create report by sex and clinic of univariate
info PROC SORT DATA wt BY sex clinic PROC
UNIVARIATE DATA wt NOPRINT BY sex clinic
VAR wt12 OUTPUT OUTunivinfo N n
Q1 p25 MEDIAN p50 Q3 P75
Dataset univinfo will have one observation
for each combination of sex and clinic.
Name of new dataset
Statistic name variable name
7
PROC PRINT DATA univinfo FORMAT sex
sexF. RUN Obs sex clinic n
p75 p50 p25 1 Male A
9 208.00 190.00 180.00 2 Male
B 16 218.25 174.75 158.25 3
Male C 29 220.50 199.50
178.00 4 Male D 11 194.00
184.50 172.00 5 Female A 6
160.50 143.50 125.00 6 Female B
9 184.00 164.50 150.00 7
Female C 6 138.50 134.25
132.50 8 Female D 6 148.50
137.50 131.00
8
PROC PRINT DATA univinfo NOOBS VAR sex
clinic n p25 p50 p75 FORMAT p25 p50 p75 6.1
FORMAT sex sexF. TITLE 'Quartiles of Weight by
Gender/Center' RUN Quartiles of Weight by
Gender/Center sex clinic N P25
P50 P75 Male A 9 180.0
190.0 208.0 Male B 16
158.3 174.8 218.3 Male C 29
178.0 199.5 220.5 Male D
11 172.0 184.5 194.0 Female A
6 125.0 143.5 160.5 Female B
9 150.0 164.5 184.0 Female
C 6 132.5 134.3 138.5 Female
D 6 131.0 137.5 148.5
9
Using ODS to Send Output to a SAS Dataset
Syntax ODS OUTPUT output-table new-data-set
Output quantile table to a dataset ODS OUTPUT
quantiles qwt PROC UNIVARIATE DATA wt VAR
wtbl wt12 RUN ODS OUTPUT CLOSE PROC PRINT
DATAqwt RUN
10
Display of Output Dataset
Obs Varname Quantile Estimate
1 wtbl 100 Max 279.30 2 wtbl 99
274.15 3 wtbl 95
246.40 4 wtbl 90 237.40 5
wtbl 75 Q3 215.15 6 wtbl 50
Median 192.65 7 wtbl 25 Q1
165.90 8 wtbl 10 141.50 9
wtbl 5 137.40 10 wtbl 1
130.25 11 wtbl 0 Min
128.50 12 wt12 100 Max 271.50 13
wt12 99 271.50 14 wt12 95
239.00 15 wt12 90
227.00 16 wt12 75 Q3 202.50 17
wt12 50 Median 180.00 18 wt12 25 Q1
153.50 19 wt12 10
133.00 20 wt12 5 130.00 21
wt12 1 123.00 22 wt12 0 Min
123.00
Would like to put side-by-side
11
DATA wtbl wt12 SET qwt if varname 'wtbl'
then output wtbl else if varname 'wt12' then
output wt12 RUN PROC DATASETS MODIFY
wtbl RENAME estimate wtbl MODIFY wt12
RENAME estimate wt12 RUN DATA all MERGE
wtbl wt12 DROP varname RUN PROC PRINT
Separate the data into 2 datasets
PROC DATASETS used for changing variable names
Put 2 datasets side-by-side
12
Obs Quantile wtbl wt12
1 100 Max 279.30 271.50 2
99 274.15 271.50 3
95 246.40 239.00 4 90
237.40 227.00 5 75 Q3
215.15 202.50 6 50 Median
192.65 180.00 7 25 Q1
165.90 153.50 8 10
141.50 133.00 9 5
137.40 130.00 10 1
130.25 123.00 11 0 Min
128.50 123.00
13
ODS OUTPUT ParameterEstimates (persistproc)
betas PROC REG DATAWT MODEL dbpchg wtchg
age sex RUN PROC REG datawt MODEL sbpchg
wtchg age sex RUN ODS OUTPUT CLOSE PROC PRINT
DATAbetas RUN
14
Display of Output Dataset - Report
Obs Dependent Variable Estimate
StdErr tValue Probt 1 dbpchg
Intercept -0.059 6.431 -0.01
0.99 2 dbpchg wtchg 0.175
0.073 2.38 0.02 3 dbpchg
age -0.101 0.112 -0.91
0.37 4 dbpchg sex -2.622
1.572 -1.67 0.10 5 sbpchg
Intercept -3.849 13.304 -0.29
0.77 6 sbpchg wtchg 0.364
0.152 2.40 0.02 7 sbpchg
age -0.042 0.231 -0.18
0.86 8 sbpchg sex -4.118
3.253 -1.27 0.21
15
Display of Output Dataset Using BY Statement
PROC PRINT VAR variable estimate stderr tvalue
probt BY dependent NOTSORTED FORMAT estimate
7.3 stderr 7.3 probt pvalue5.2 Dependentdbpchg
Obs Variable Estimate StdErr
tValue Probt 1 Intercept -0.059
6.431 -0.01 0.99 2 wtchg
0.175 0.073 2.38 0.02 3 age
-0.101 0.112 -0.91 0.37
4 sex -2.622 1.572 -1.67
0.10 Dependentsbpchg Obs Variable
Estimate StdErr tValue Probt 5
Intercept -3.849 13.304 -0.29
0.77 6 wtchg 0.364 0.152
2.40 0.02 7 age -0.042
0.231 -0.18 0.86 8 sex
-4.118 3.253 -1.27 0.21
16
PROC RANK

• Used to divide observations into equal size
  categories based on values of a variable
• Creates a new variable containing the categories
• New variable is added to the dataset or to a new
  dataset
• Example Divide weight change into 5 equal
  categories (Quinitiles)

17
PROC RANK SYNTAX
PROC RANK DATA dataset OUT
outdataset GROUPS of categories
VAR varname RANKS newvarname Most of the time
you can set OUT to be the same dataset specified
in DATA. PROC RANK writes no output
18
PROGRAM 15 LIBNAME class C\SAS_Files' DATA
wtchol SET class.tomhsp (KEEPptid clinic sex
wtbl wt12 cholbl chol12) wtchg wt12
- wtbl cholchg chol12 - cholbl RUN This
PROC will add a new variable to dataset which is
the tertile of weight change. The new variable
will be 0,1,or 2 PROC RANK DATA wtchol
GROUPS3 OUT wtchol VAR wtchg RANKS
twtchg
Name of new variable
19
PARTIAL LOG 8 DATA wtchol 9 SET
class.tomhsp (KEEPptid clinic sex wtbl wt12
cholbl chol12) 10 wtchg wt12 - wtbl 11
cholchg chol12 - cholbl 12 RUN NOTE There
were 100 observations read from the data set
CLASS.TOMHSP. NOTE The data set WORK.WTCHOL has
100 observations and 9 variables. PROC RANK
DATA wtchol GROUPS3 OUT wtchol 20 VAR
wtchg RANKS twtchg 21 RUN NOTE The data
set WORK.WTCHOL has 100 observations and 10
variables.
20
PROC FREQ DATA wtchol TABLES twtchg
RUN OUTPUT Rank for Variable wtchg
Cumulative
Cumulative twtchg Frequency Percent
Frequency Percent ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ
ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ 0
31 33.70 31 33.70 1
30 32.61 61
66.30 2 31 33.70
92 100.00 Frequency Missing 8
21
PROC PRINT DATA wtchol (obs20) VAR ptid
wtchg twtchg TITLE 'Partial Listing of Datset
wtchol with new variable added' RUN Parti
al Listing of Datset wtchol with new variable
added Obs PTID wtchg twtchg 1
A00083 -12.00 1 2 A00301
. . 3 A00312 -9.50 1
4 A00354 -21.00 0 5 A00400
. . 6 A00504 -9.25 1
7 A00608 . . 8 A00720
-18.50 0 9 A00762 -5.25
2 10 A00811 -6.75 1
22
PROC MEANS N MEAN MIN MAX MAXDEC2 VAR cholchg
wtchg CLASS twtchg TITLE 'Mean Cholesterol
Change by Tertile of Weight Change' RUN
23
Mean Cholesterol Change by Tertile of Weight
Change The MEANS Procedure Rank for
Variable N wtchg Obs Variable N
Mean Minimum Maximum -----------
--------------------------------------------------
------------- 0 31 cholchg 30
-13.43 -55.00 47.00
wtchg 31 -22.51 -36.50
-14.30 1 30 cholchg 30
-4.70 -37.00 26.00
wtchg 30 -10.21
-14.00 -6.80 2 31 cholchg
31 -0.74 -52.00 45.00
wtchg 31 -1.82
-6.50 13.00 ----------------------------
---------------------------------------------- Co
uld graph this data in an x-y plot (3 points)
Cutpoints for tertiles