delete outliers

[sasphd]

I want simply delete outliers 0.5% at each tail. How I can do that???

[art297]

Take a look at the ideas (and solutions) presented in this thread: https://communities.sas.com/t5/SAS-Statistical-Procedures/How-to-remove-OUTLIERS-in-Base-SAS/td-p/90...

HTH,

Art, CEO, AnalystFinder.com

[sasphd]

I did not suceed to do it Can you give me a sample

[mkeintz]

What did you do?  What did you expect it to do?  What did you get?

[art297]

Here is an example:

data have;
  input Cost;
  cards;
11326.210369
12486.48938
12461.600695
17078.181683
20830.827535
21301.654723
19246.458983
17233.110851
20086.991101
18998.225999
18755.580299
18363.870699
18084.353599
22674.720499
16680.130049
15838.031999
19773.713649
16864.323849
17499.784599
16918.186999
18383.525249
20307.965549
18135.335899
16504.779198
22346.398306
18484.330644
18335.447644
19685.665094
19591.907594
18702.066744
18938.583444
17878.801744
23398.002544
18101.282344
18554.513094
18812.566344
21834.429194
19587.505132
16619.15172
;

proc rank data=have out=ranks (drop=cost)  groups=100;
  var cost;
  ranks cost_rank;
run;

data want;
  set have;
  set ranks;
  if 6 <= cost_rank <= 95;
run;

HTH,

Art, CEO, AnalystFinder.com

[atzamis]

Hi,

try this

proc rank data=have out=ranks(where=(6 <= cost_rank <= 95 ))   groups=100;
  var cost;
  ranks cost_rank;
run;