The first SAS PROCs

I was recently following a post where a user asked what the earliest SAS PROCs (procedures) were. Since I started using SAS in 1970, I knew I could find a copy of the old documentation that we used at that time. This “user’s guide” was something that you printed out on an old line printer on regular computer paper.

I did find the guide. And the list of procedures in SAS at that time brought back some good memories. By the time I found the old guide, Chris Hemedinger had already responded to the user on Facebook, but hey, he doesn’t go back as far as I do (most people don’t!). So, here is the list of PROCs documented in the 1970 SAS User’s Guide:

  • PROC REGR - regression
  • PROC ANOVA - analysis of variance
  • PROC FACTOR - factor analysis
  • PROC AOV - alternate analysis of variance (treats missing values as zero; not as efficient as ANOVA)
  • PROC LATTICE - analysis of variance with three types of analysis (square balanced, square unbalanced, rectangular)
  • PROC DUNCAN - Duncan’s multiple range test
  • PROC PLAN - randomized plans
  • PROC CORR - bivariate statistics
  • PROC SORT
  • PROC MEANS
  • PROC NESTED - nested analysis of variance
  • PROC PRINT
  • PROC FREQ

Here is Chris Hemedinger's response to the user. 

Anyone go back farther than that???

 

 

 

 

 

tags: chris hemedinger, PROCs, sas, SAS PROC, SAS User's Guide

12 Comments

  1. Posted July 12, 2012 at 9:22 am | Permalink

    I only go back as far as 1975. But I do remember that I used each and every one of these PROCS!

  2. Jason Secosky
    Posted July 17, 2012 at 9:57 am | Permalink

    While not technically a PROC, I'm sure the DATA step would be worthy of being on this list too!

  3. Archer Gravely
    Posted July 17, 2012 at 5:04 pm | Permalink

    I stilI have my 1979 SAS manual and it cost $9.95 in the Va Tech Bookstore...That was pricey for a grad student at the time.

  4. Posted July 17, 2012 at 5:16 pm | Permalink

    I was using FORTRAN (WATFOR) in 1970. After finding out about SAS in the late 70's I never looked back, but I was sure glad when we did away with punch cards and paper tape! When I found out about JMP (version 2) my SAS use dropped dramatically but it did not go to zero.

  5. Erik Tilanus
    Posted July 17, 2012 at 5:17 pm | Permalink

    Kathy,

    Great! It makes me feel like a newbie!
    The first manual I put my hands on (and made me recommend to my boss at KLM to take a SAS license) was the 1976 manual. The oldest documentation I still have is the 1978 edition of the SAS Introductory Guide by Jane Helwig.

  6. Don Henderson
    Posted July 17, 2012 at 5:25 pm | Permalink

    Jason, way back when, I vaguely recall Tony Barr referring to the DATA step as PROC DATA. He said that conceptually a DATA step was just a special kind of a PROC.

  7. Gabe
    Posted July 17, 2012 at 5:38 pm | Permalink

    What was the functionality of PROC SORT? Did the other procs allow for by-group processing, or was PROC SORT solely for data step processing?

    • Kathy Council Kathy Council
      Posted July 18, 2012 at 10:58 am | Permalink

      You could sort the data with PROC SORT, then use the sorted data in procs. Is that what you are asking?

  8. Robert H. Miller
    Posted July 18, 2012 at 10:12 am | Permalink

    Seeing your note concerning history of SAS Procs brings back memories to me, although exact dates are beyond recall.

    I spent my career with the Agricultural Research Service at Beltsville Maryland (after a Ph.D. at NCSU in 1962). In those days availability of statistical software was spotty. In the 60's we developed the Proc Nested program for hierarchical analysis of variance (written by Merrill Swanson in Fortran I believe). During the time Drs. Goodnight and Sall were getting SAS Inc. off the ground, there was considerable cooperation with Beltsville. One aspect was our contribution of Proc Nested. I also remember that Dr. Tommye Cooper who ran the Beltsville computer Lab assisted with aspects of testing and development. We were overjoyed with the opportunity to use SAS for statistical work instead of employing half a dozen different programs which didn't communicate with each other, such as Nested and Walt Harvey's least-squares program. Many Beltsville scientists benefitted through the use of SAS. Sorry I can't remember more precise dates.

    Robert H. Miller
    Columbia MD

  9. Richard Wright
    Posted July 19, 2012 at 10:46 am | Permalink

    I cut my teeth on step-wise autoregression in mid 70's as an undergrad. Prof tossed me a book and said something to the effect of "Here's something called SAS. Go learn it."

    I seem to remember PROC BANNER being in the earliest releases, although I'm not sure it was documented.

    Speaking of JMP - somewhere at home, I still have the beta of JMP on a 3.5" (?) floppy disk. I'm using it more and more, although frequently have it and SAS up at the same time and go back and forth between the two.

  10. Joe O'Hop
    Posted July 20, 2012 at 1:26 pm | Permalink

    I first used SAS during graduate school at UGA in the mid-70's. Loading the card reader with boxes of cards, a little pre-processing in Fortran, then on to SAS for analyses. I recall many mornings seeing the sunrise after nights spent at the card punch machine and reader. Punching cards, dropping cards, re-ordering cards, reading cards, fixing cards, etc. I'm very glad that these are just memories nowadays. My friends and I sometimes speculated back then about the potential for development of "Proc Thesis" and "Proc Dissertation". They might have been best-sellers.....

  11. Nina
    Posted August 10, 2012 at 8:35 am | Permalink

    Wow you've been using SAS since before I was born - didn't realise it has been around that long. I started using it in 2007 during my MSC Statistics.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <p> <pre lang="" line="" escaped=""> <q cite=""> <strike> <strong>