cosd and sind

[McCoder]

I'm using Fortran g95 to compile my code and receive a 'undefined reference' error to sind and cosd.

      real glong,glat,utsec,tau, season, cos tau, alpha, phi
      real pi,c,theta, Sx, Sz, Nx, Nz, hgraze,Re,al,aet
      real sind, cosd
      integer*4 day
      data alpha/.4089/                  
      data pi/3.14159265/
      data Re/6.37E3/                    
      c = 2.*pi/360.

      al = 278.895 + 0.985647*day
      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -
     % 12.7*sind(4*al) - 429.1*cosd(al) - 2.1*cosd(2*al) +
     % 19.3*cosd(3*al)
      season = 2.*pi*(day-173.21+((utsec-aet)/86400.))/365.25
      phi = c*(glong + (utsec-aet)/240.)
      theta = c*(90. - glat)

What is the best way to define cosd and sind? Should I use parameter?

[evansbw]

which compiler is this for (Intel, I assume)

I have not used Fortran in many years, but recently used Intel, and I'm wondering if "cosd" and "sind" are library functions - have your checked (eg, are they "reserved" symbols)? If so, perhaps try changing the variable identifier to something like mysined and mycosd.

After re-reading the code, I noticed the variables cosd, and sined, are not initialized. Is this important?? Different languages auto-initialize differently (some do not!)

[clowenstein]

which compiler is this for (Intel, I assume)

I have not used Fortran in many years, but recently used Intel, and I'm wondering if "cosd" and "sind" are library functions - have your checked (eg, are they "reserved" symbols)? If so, perhaps try changing the variable identifier to something like mysined and mycosd.

After re-reading the code, I noticed the variables cosd, and sined, are not initialized. Is this important?? Different languages auto-initialize differently (some do not!)

Back in the olden days these functions were named dcos and dsin. Who wrote the code that you are trying to compile?

[drl]

Hi.

At a quick glance, this looks like non-standard functions for sin/cos with degree arguments.

If so, you could Google for fortran sind cosd degree function and you will find some references, for example:

real function sind(x)
sind=sin(x/180.0*3.141592)
return
end

from WMS FORTRAN: what are SIND, COSD, etc? - Rhinocerus

Best wishes ... cheers, drl

[McCoder]

Thanks for the help, all!

Just as a note, I'm using gfortran, not g95. I'm simply modifying this code for a project. The original code was written in 1985 (I wasn't even born then).

My new code looks like this:

      real glong,glat,utsec,tau, season, cos tau, alpha, phi
      real pi,c,theta, Sx, Sz, Nx, Nz, hgraze,Re,al,aet,sind,cosd,x
      integer*4 day
      parameter (alpha=.4089)                  
      parameter (pi=3.14159265)
      parameter ( Re=6.37E3)                   
      c = 2.*pi/360.
      sind = sin(x/180.0*pi)
      cosd = cos(x*pi/180.0)

      al = 278.895 + 0.985647*day
      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -
     % 12.7*sind(4*al) - 429.1*cosd(al) - 2.1*cosd(2*al) +
     % 19.3*cosd(3*al)
      season = 2.*pi*(day-173.21+((utsec-aet)/86400.))/365.25
      phi = c*(glong + (utsec-aet)/240.)
      theta = c*(90. - glat)

However, now I receive the error:

      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -       
     1
Error: Unclassifiable statement at (1)

This doesn't make sense to me because I assign REAL aet.
Suggestions? Thanks again!

[drl]

Hi.

      parameter ( Re=6.37E3)                   
      sind(x) = sin(x/180.0*pi)
      cosd(x) = cos(x*pi/180.0)
      c = 2.*pi/360.

cheers, drl

[ImChipBrown]

Thanks for the help, all!

Just as a note, I'm using gfortran, not g95. I'm simply modifying this code for a project. The original code was written in 1985 (I wasn't even born then).

My new code looks like this:

      real glong,glat,utsec,tau, season, cos tau, alpha, phi
      real pi,c,theta, Sx, Sz, Nx, Nz, hgraze,Re,al,aet,sind,cosd,x
      integer*4 day
      parameter (alpha=.4089)                  
      parameter (pi=3.14159265)
      parameter ( Re=6.37E3)                   
      c = 2.*pi/360.
      sind = sin(x/180.0*pi)
      cosd = cos(x*pi/180.0)

      al = 278.895 + 0.985647*day
      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -
     % 12.7*sind(4*al) - 429.1*cosd(al) - 2.1*cosd(2*al) +
     % 19.3*cosd(3*al)
      season = 2.*pi*(day-173.21+((utsec-aet)/86400.))/365.25
      phi = c*(glong + (utsec-aet)/240.)
      theta = c*(90. - glat)

However, now I receive the error:

      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -       
     1
Error: Unclassifiable statement at (1)

This doesn't make sense to me because I assign REAL aet.
Suggestions? Thanks again!

Just a guess. Looks like the error occurs at the line break.

[clowenstein]

Good catch.

If that doesn't fix everything, I would try not using % as the continuation marker in column 6. Maybe simple integers like 1 and 2, as has been Fortran standard since the late 1950's.

[drl]

Hi.

The original problem was due to the use of non-standard trig functions sind & cosd.

The OP added statement functions to handle those. However, the syntax was wrong, and the placement was wrong (statement functions cannot appear after executable statements).

Correcting these with the 4 lines as noted in post # 6 results in a code that can compile without syntax errors. The following script lists the environment, the source code, and results from compiling with 3 compilers:

#!/usr/bin/env bash

# @(#) s1	Demonstrate Fortran-77 statement function syntax, placement.

# Infrastructure details, environment, commands for forum posts. 
# Uncomment export command to run script as external user.
# export PATH="/usr/local/bin:/usr/bin:/bin"
set +o nounset
pe() { for i;do printf "%s" "$i";done; printf "\n"; }
pl() { pe;pe "-----" ;pe "$*"; }
LC_ALL=C ; LANG=C ; export LC_ALL LANG
pe ; pe "Environment: LC_ALL = $LC_ALL, LANG = $LANG"
pe "(Versions displayed with local utility \"version\")"
c=$( ps | grep $$ | awk '{print $NF}' )
version >/dev/null 2>&1 && s=$(_eat $0 $1) || s=""
[ "$c" = "$s" ] && p="$s" || p="$c"
version >/dev/null 2>&1 && version "=o" $p printf specimen gfortran g95 ifort
set -o nounset
pe

FILE=${1-t1.f}

# Display sample of data file, with head & tail as a last resort.
pe " || start [ first:middle:last ]"
specimen 10:0:10 $FILE \
|| { pe "(head/tail)"; head -n 5 $FILE; pe " ||"; tail -n 5 $FILE; }
pe " || end"

pl " Results, gfortran:"
gfortran $FILE
ls -lgG a.out

pl " Results, g95:"
g95 $FILE
ls -lgG a.out

pl " Results, ifort (Intel):"
ifort $FILE
ls -lgG a.out

exit 0

producing:

% ./s1

Environment: LC_ALL = C, LANG = C
(Versions displayed with local utility "version")
OS, ker|rel, machine: Linux, 2.6.26-2-amd64, x86_64
Distribution        : Debian GNU/Linux 5.0 
GNU bash 3.2.39
printf - is a shell builtin [bash]
specimen (local) 1.17
gfortran GNU Fortran (Debian 4.3.2-1.1) 4.3.2
G95 (GCC 4.0.3 (g95 0.92!) Jun 24 2009)
ifort (IFORT) 11.1 20090827

 || start [ first:middle:last ]
Whole: 10:0:10 of 19 lines in file "t1.f"
      real glong,glat,utsec,tau, season, cos tau, alpha, phi
      real pi,c,theta, Sx, Sz, Nx, Nz, hgraze,Re,al,aet,sind,cosd,x
      integer*4 day
      parameter (alpha=.4089)                  
      parameter (pi=3.14159265)
      parameter ( Re=6.37E3)                   
      sind(x) = sin(x/180.0*pi)
      cosd(x) = cos(x*pi/180.0)
      c = 2.*pi/360.

      al = 278.895 + 0.985647*day
      aet = -105.7*sind(al) + 596.2*sind(2*al) + 4.4*sind(3*al) -
     % 12.7*sind(4*al) - 429.1*cosd(al) - 2.1*cosd(2*al) +
     % 19.3*cosd(3*al)
      season = 2.*pi*(day-173.21+((utsec-aet)/86400.))/365.25
      phi = c*(glong + (utsec-aet)/240.)
      theta = c*(90. - glat)
      end

 || end

-----
 Results, gfortran:
-rwxr-xr-x 1 10899 Jun 29 09:45 a.out

-----
 Results, g95:
-rwxr-xr-x 1 578949 Jun 29 09:45 a.out

-----
 Results, ifort (Intel):
-rwxr-xr-x 1 723014 Jun 29 09:45 a.out

cheers, drl