Variable defined as both character and numeric

[Barkamih]

 Hi

 I have a problem when I run this code in order to combine two datasets together by using this code, Calving_ease is the issue, this code can present only the number of calving_ease of the first dataset, but the numbers of calving_ease  in the second dataset are all missing.

regards 

data  dataall;

set data48 data910(rename=(calving_ease=calving_easenum));

calving_ease = put(calving_ease, 7.);

drop calving_easenum;

run;

[RW9]

In dataset date48 you have a variable called calving_ease, in data910 you also have this, however you rename it to calving_easenum  Now when these two datasets are set together the output creates a variable called calving_ease which holds the data from the first table, and fills in empties from the second, as that variable does not exist.  This is the logic you have presented below, and hence that is what is achieved at the end.  I cannot guess what you are trying to do, nor can I understand why your thread topic is char vs num but that is not what the question you present is.  A quick check on the guidance when presenting questions would help, present test data in the form of a datastep, show what you want out at the end.  I am going to guess that in one dataset calving_ease is character, and the other is numeric?  If so convert one or the other before setting.

[Barkamih]

yeas 

your  guess is right because I had this issue I did this code, the normal code for  combining  data is this

data all;

set data1 data2;

run;

 but I have an error with the code as the calvng_ease defined as both character and numeric. 

so I tried to do it by the previous code that I posted,  but I got missing data for the second dataset 

[novinosrin]

@Barkamih you need to convert one of the variables to either character or numeric to make it consistent for successful append

[Barkamih]

could you please provide an example code, how to convert it ?

[novinosrin · Accepted solution]

I made a small correction to your code:

assuming  calving_ease in 48 is character and 910 needs to become character

Try and let me know

data  dataall;

set data48 data910(in=b rename=(calving_ease=calving_easenum));

if b then calving_ease = put(calving_easenum, 7.);/*changed here to calving_easenum*/

drop calving_easenum;

run;

[Barkamih]

calving_ease of the second dataset are all missing by using this code

[novinosrin]

Did you notice my edit in the previous message. Anyway, here is a demo

data have;
input var $;
datalines;
1
2
3
;

data have1;
input var;
datalines;
1
2
3
;

data want;
set have have1(in=b rename=(var=_var));
if b then var=put(_var,1.);
drop _var;
run;

[novinosrin]

Your approach is bang on correct. I think you are having a very minor problem:

instead of

calving_ease = put(calving_ease, 7.);

I changed the same to 

calving_ease = put(calving_easenum, 7.);

 plus instead of 

set data48 data910(rename=(calving_ease=calving_easenum));

I changed this to 

set data48 data910(in=b rename=(calving_ease=calving_easenum));

[Barkamih]

thank you much, I appreciate that, you are done a very good jobe with me, and I'm sorry if I took you time 

best regards 

thanks again 

Barkamih 

[novinosrin]

@Barkamih  Hahaha nothing to be apologetic. We all learn and I am sure you share your knowledge too with tohers. Knowledge is meant to be shared and learning good things should a right. Take care and have a nice day!