Help using Base SAS procedures

How to calculate mode for length

Reply
Super Contributor
Posts: 400

How to calculate mode for length

I have a column that has text in each row.  I was assigned to find the mode of this column by my manager.  I have never heard of mode in this respects so i'm not sure how I would calculate it in sas.  Any help would be great appreciated.

Super User
Posts: 19,851

Re: How to calculate mode for length

Posted in reply to jerry898969

proc freq and sort descending will get you the most freq. I think there's an option to set it to sort descending.

Super Contributor
Posts: 400

Re: How to calculate mode for length

Reeza,

Thank you for your help I will try that out.  Is there a proc that can give me mean, median and mode all at once? If not what would be the best approach to calcualte mean and median?

Thank you so much

Super User
Posts: 19,851

Re: How to calculate mode for length

Posted in reply to jerry898969

For a text variable?

I don't know how mean or median for those would be applicable.

Mode makes some sort of sense to find the most frequent comment or word.

For numeric variables, use proc means/summary/univariate.

Super Contributor
Posts: 400

Re: How to calculate mode for length

Sorry I didn't explain it better.  It is the length of the text field I need these statistics for.  I will try those procs thanks.

Super User
Posts: 11,343

Re: How to calculate mode for length

Posted in reply to jerry898969

Data want;

     set have;

     VarLength=length(variablename of interest);

run;

proc means data=want mean median mode;

var Varlength;

run;

Occasional Contributor
Posts: 5

Re: How to calculate mode for length

Posted in reply to jerry898969

In three ways you can calculate mean, mode and median of numeric variable.

1) Proc menas data=<dataset_name> mean mode median;

    var <variable_names>;

     run;

2) Proc summary data=<dataset_name> print mean mode median;

    var <variable_names>;

     run;

3) ODS select basicmeasures;

     Proc univariate data=<dataset_name>;

     var <variable_names>; 

     run;

Check it out.

-Prit

Ask a Question
Discussion stats
  • 6 replies
  • 274 views
  • 0 likes
  • 4 in conversation