Proc transpose not keeping all decimal values

[mahi2pal]

data Abun_dw20d_summary ;       informat bins $30.; format bins $30.; length bins $30.;
   input bins $ CM1DW20D CM3DW20D CW1DW20D ODWD W1DW4D;
   datalines;         
Acinetobacter 2477.123456789 195  163 45 77
Actinoalloteichus 2431 220  198 69.123456789 6774
Bacillus 2456 173  155 78 785
Bletilla 2412 135  116.123456789 45 123;   
run;


proc transpose data=Abun_dw20d_summary out=bint;
    var _all_;
    by bins;
run;

My decimals are getting truncated in output dataset, can anyone help me on this? 

[data_null__]

---

@mahi2pal wrote:

data Abun_dw20d_summary ;       informat bins $30.; format bins $30.; length bins $30.;
   input bins $ CM1DW20D CM3DW20D CW1DW20D ODWD W1DW4D;
   datalines;         
Acinetobacter 2477.123456789 195  163 45 77
Actinoalloteichus 2431 220  198 69.123456789 6774
Bacillus 2456 173  155 78 785
Bletilla 2412 135  116.123456789 45 123;   
run;


proc transpose data=Abun_dw20d_summary out=bint;
    var _all_;
    by bins;
run;

My decimals are getting truncated in output dataset, can anyone help me on this? 

---

Did you notice this?

NOTE: Numeric variables in the input data set will be converted to character in the output data set.

SAS used a default format I think BEST12.

If you want more decimal places use a different format.

[Cynthia_sas · Accepted solution]

Hi:

  Providing a format seems to take care of the issue for me:

Hope this helps,

cynthia

[mahi2pal]

Where do give the format?

[Cynthia_sas]

It is highlighted in yellow in the screen shot in my previous posting.
Cynthia

[Tom]

1) Don't include the character variables in the VAR statement and PROC TRANSPOSE will not convert your numeric variables to character.

2) If you must include the character variables then attach a format to the numeric variables that is long enough.

Example:

data have ;       
  length bins $30.;
  input bins CM1DW20D CM3DW20D ;
datalines;         
Acinetobacter 2477.123456789 195 
Bletilla 2412 116.123456789
;
proc print width=min; title 'HAVE';
  format _numeric_ best32.;
run;

proc transpose data=have out=want1;
  by bins;
run;
proc print width=min; title 'WANT1';
  format _numeric_ best32.;
run;
proc transpse data=have out=want2;
  by bins;
  var _all_;
  format _numeric_ best32.;
run;
proc print width=min; title 'WANT2'; 
run;
title;

HAVE
Obs    bins                   CM1DW20D         CM3DW20D

 1     Acinetobacter    2477.123456789              195
 2     Bletilla                   2412    116.123456789

WANT1
Obs    bins              _NAME_               COL1

 1     Acinetobacter    CM1DW20D    2477.123456789
 2     Acinetobacter    CM3DW20D               195
 3     Bletilla         CM1DW20D              2412
 4     Bletilla         CM3DW20D     116.123456789

WANT2
Obs    bins             _NAME_                    COL1

 1     Acinetobacter    bins        Acinetobacter
 2     Acinetobacter    CM1DW20D                      2477.123456789
 3     Acinetobacter    CM3DW20D                                 195
 4     Bletilla         bins        Bletilla
 5     Bletilla         CM1DW20D                                2412
 6     Bletilla         CM3DW20D                       116.123456789