Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uunet!samsung!zaphod.mps.ohio-state.edu!mips!sgi!shinobu!odin!dinkum!calvin
From: calvin@dinkum.sgi.com (Calvin H. Vu)
Newsgroups: comp.lang.fortran
Subject: Re: problem with "pe" format
Message-ID: <8474@odin.corp.sgi.com>
Date: 4 Jun 90 22:10:13 GMT
References: <90155.122858GL4@psuvm.psu.edu>
Sender: news@odin.corp.sgi.com
Lines: 39

In article <90155.122858GL4@psuvm.psu.edu> GL4@psuvm.psu.edu writes:

>---------------------------------------------
>       x=1.11
>       fs=10.0
>       write(*,32) x,fs
>32     format(1pe13.3,f10.1)
>       write(*,*) fs
>       end
>---------------------------------------------
>output:
>       1.111E+00     100.0
>      10.00000
>---------------------------------------------

> Does anyone know why the value of "fs" is multiplied by a factor of
> ten? I've tried this on a Sun-3, a Gould 9005 running UTX/32 2.1A,
> and an IBM mainframe, and gotten the same results.

The P Editing causes every real number following it (until another P 
editing if any) to be multiplied by 10**k where k is called the scale
factor and is the number in front of the P.  A less confusing way to
write this is:
32     format(1p,e13.3,f10.1)

Without the scale factor specified, i.e. if the format had been:
32     format(e13.3,f10.1)

the result would have been:
    0.111E+01      10.0
    10.00000    

If you note the difference between the two results you can figure out the
effects of the P edit descriptor.
I'm a little lazy to type all the details so maybe you do need to RTFM :-).
--------------------------------------------------------------------------
Calvin H. Vu			   | "We are each of us angels with
Silicon Graphics Computer Systems  | only one wing.  And we can only
calvin@sgi.com   (415) 962-3679	   | fly embracing each other."