importing data listed in billions of dollars

[Damon1]

Hi Everyone, 

Been learning for about a week. So, i have a data set that follows the following pattern:

rank (numeric), company ($), country ($), sales (num), profits (num), assets (num), market value (num)

the data under the last 4 numeric variables are presented in billions of dollars, as "$1.5B" or "$200.1B", and vary in length 

Im using the following code:

DATA bigcos;
INFILE 'C:\Users\dlobsien\Documents\bigcompanies.txt';
INPUT ranking company_name $26. +1 country $18. sales_in_billions :dollar10. profits_in_billions :dollar10. assets_in_billions :dollar10. +1 profits_in_billions :dollar10.;
title companies data;
RUN;

——————————-

PROC PRINT data=bigcos (obs=15);
Run;

the first three variables go in fine, but it wont read anything after that with the Bs in the way (i assume that's the issue). whenever i add a line that says "compress(var, 'B')", after the input lines it gives me 0 observations—maybe I'm entering the syntax wrong or putting it in the wrong spot. 

I've also included the data set in case that is useful.

Any help would be greatly appreciated.

[Tom]

You mentioned that the values vary in length, but it does not look that way when I viewed your attached text file using this websites viewer tool.  The values of a given field are always in the same location on the line.

Why not just skip that column in your INPUT statement. Looks like you already know how to skip a column since you have included the +1 cursor movement command in your current INPUT statement.

If the value is not always a B then read it as another character variable and then you can write a program to adjust the magnitude of the number read based on what letter was appended.

Hint: You don't want to use the : modifier on your informats.  That will switch your input statement from formatted mode to list mode.

[Damon1]

yeah, I'm not actually sure how to use the +1. i added on a whim it because i noticed it in a textbook, and added it one at a time because I think made the character variables read.

the data under the last 4 numeric variables vary from as short as $4.9B to as long as $2,265.8B, and each observation for those variables ends in "B"

[Tom]

Include the leading spaces when counting how many characters the field takes. So if the field before the one with B in it ends in column 20 and the B is in column 30 then use COMMA9. informat to read the 9 characters in columns 21 to 29.
... @21 number comma9. letter $1. ...

[Reeza · Accepted solution]

https://communities.sas.com/t5/SAS-Programming/How-do-I-properly-use-PROC-IMPORT-for-a-multi-spaced-...

Classmates?

You need to write your own informat preferably but you could also just read it as a character, remove the B and then convert it to billions. 

https://www.lexjansen.com/pharmasug/2005/posters/po06.pdf

[Damon1]

probably—its an online class so i couldnt tell you lmao.

i imported it as a character, then compressed out the $ and B then made a new data set and wrote varname=import(oldvar, best.)

and that worked well enough