- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Dear community,
In the SAS Statistical Business Analysis Using SAS 9 certification guide, the chapter presenting the Chi-Square Tests mentions the opportunity to use the variable worth to order predictor variables by importance:
However, I can't find this calculation in the PROC FREQ output when I use the CHISQ option in the TABLES statement. I have tried to find out if there is another specific option but I haven't found it.
For example, the output for their example is:
proc freq data=cert.ames300;
tables bonus*High_Kitchen_Quality / chisq;
run;
Do you know if SAS can automatically compute this value? Or do we need to go through a data step by computing something like -2*log10(p_pchi)?
Thank you very much for your help.
Best,
- Tags:
- chisq
Accepted Solutions
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
If you go to https://support.sas.com/en/search.html?q=variable%20worth&fq=releasesystem%3A9.4 and search for variable worth, the only real links you find are for Enterprise Miner.
Paige Miller
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
If you go to https://support.sas.com/en/search.html?q=variable%20worth&fq=releasesystem%3A9.4 and search for variable worth, the only real links you find are for Enterprise Miner.
Paige Miller
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Thank you @PaigeMiller !
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Note that a plot of -log(p) is provided by PROC MULTTEST when the PLOTS=MANHATTAN option is specified. See the Getting Started example and try it with that option.
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Thank you @StatDave !