Dear Sir,
I wish to split my sample to 3 groups based on total assets (variable:at), ie to small firm, medium firm and big firm. this means i need to calculate the 1/3 quantile, 2/3 quantile for the variable.
How do i write that in the program?
also: if i wish to split to 4 parts, i suppose Q1, median, Q3 that is generated from the normal univariate procedure would be fine right?
Thanks a lot!!
hi ... quantiles and the median are based on counts, not magnitude of numbers so that approach will not categorize small/ medium/large
for example, if I looked at this data, I might say that 1 is small, 100 is medium, 1000 is large ... but the quantiles and median don't say that
data test;
input assets @@;
datalines;
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 10 10 1000
;
proc univariate data=test;
ods select quantiles;
run;
Variable: assets
Quantiles (Definition 5)
Quantile Estimate
100% Max 1000
99% 1000
95% 100
90% 100
75% Q3 1
50% Median 1
25% Q1 1
10% 1
5% 1
1% 1
0% Min 1
it is okay, i just want the count as my data size is very big, up to 20,000 observations. i just need to split them to 3 groups.
can you guide me in writing the codes to 3 parts?
hi ... specify the number of groups in GROUPS= ... the lowest group is always 0
with three groups, the value of ASSET_GROUP will be 0, 1, 2
proc rank data=test out=test_group groups=3 ties=low;
var assets;
ranks asset_group;
run;
(you can look at online help for an explanation of TIES=)
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.