compilation problem : abstract type and deferred procedure

As an exercise ( I learn Fortran up to F2003 - with great pleasure by the way ) I tried to write a small program using an abstract type and a deferred function ( code below ). As it is: Using ifort I got an Internal Compiler Error ( and posted something at the Intel compiler forum ) Using gfortran I have an error: myPoint = point2d(2.3, 4.2) 1 Error: Invalid character in name at (1) Replacing this line by : myPoint%x = 2.3 ; myPoint%y = 4.2 everything works. Does someone here see a problem with the code ? Marc module pointModule implicit n

cos intrinsic and double precision

Hello, I am a newcomer to Fortran and I have the following question. Does the cos intrinsic in fortran returns a double precision number o a single precision one. Reading the documentation I am lead to believe the latter, but then how do I calculate trigonometric functions in double precision. Thank you in advance, Makis ps. I am using g77 from the gcc4.3.6 collection.

Format reading

Ok guys I made a mistake in the reading language: Should be IF (HHMM1 .EQ. HHMM2) THEN DO SOMETHING END IF Now my question would go further than that. In fortran77 how do you deal with other formats, specially Excel, CSV, etc.? Thanks for your coolaboration

A plug for Code::Blocks

http://www.codeblocks.org/ I've been trying this out (on Windows, with gcc/gfortran) and I'm pretty impressed. It is a free open-source cross-platform IDE, with a high degree of configurability, but intelligent defaults. The project I've been testing with has about 10 modules created by different files, with multiple dependencies. It took me a bit of hunting around, but I found that the compilation order can be controlled by setting priority levels on the source files. I've just scratched the surface of what's available, but already I have something I can use. Another big pl

Effect of SPAM on number of active contributors

I really appreciated the give-and-take of the postings trying (and succeeding) to help me in converting the old (1978) CDC Fortan IV program. It took this war-story telling to get 88 responses. But how few the author names (aliases) that cropped up! But the non-news-service feed (Google Forum)s access to this Usenet Forum provides a 20-to-1 SPAM to Fortran posting ratio. I actually counted a 99-to-1 raion one single horrible day. I am now certain that SPAM is killing the active membership of this Forum. Some days are better than others. Wouldn't a moderated forum but a better

intent(in) / intent(out)

Just to see what happens, I tried the following: program test_intent implicit none integer :: var var = 1 call test_intent( in=var, out=var ) print*,var ! contains ! subroutine test_intent(in, out) implicit none integer, intent(in) :: in integer, intent(out) :: out ! Initialize out. out = 0 ! Calculate something. out = in*5 end subroutine test_intent end program test_intent !--------------------------------------- the output is not 5, but 0. The reason for having the input and output in one variable is simply

generic lexicographical sort in fortran

is it possible to write a sort routine with variable number of arguments and without type restrictions on the arguments plus an ascending/descending flag? Thanks.

Web site and Workshops

Hello, I've reworked my web site. It's now www.daniellnagle.com The old erols site should be ignored. I haven't kept it for a couple of years. (I'm working on my dissertation.) I've got two new workshops. One is on coarrays, the other is on using CUDA from Fortran. Please contact me if you have any interest. Further details are on the web site. -- Cheers! Dan Nagle

Re: OT: Punched cards -- was Re: non portable - a problem with direct access

Roland Hutchinson wrote: > On Sun, 29 Aug 2010 19:18:40 -0400, Joe Morris wrote: > >> "Peter Flass" <Peter_Flass@Yahoo.com> wrote: >>> dpb wrote: >>>> Ahem A Rivet's Shot wrote: >> >>>>> I know FORTRAN is old, but *that* old ? >>>> >>>> Now you know what the advantage Caesar had over the Gauls was based >>>> on... :) >> >>> The Gauls used Algol? >> >> Probably using the compiler (for IBSYS) written at Grenoble. (Anyone >> else recall Gralgol?) >> >> Side question: does anyone here remember enough of their high-school >> Latin to come up with a plausible Latin expression

Problem with allocate

I recently met a problem with allocate. The code runs smoothly all the way till the following segment: ALLOCATE(E(nlength),STAT=allo_stat) IF(MemoryError('E',error)) GOTO 9999 E = 0.0d0 I believe the error is due to allocate a big array and the system does not have sufficient memory. MemoryError is a function for reporting this error. However, instead of getting meaningful error message, the system crashed instead. Can somebody tell me why and how to present such situations? I cannot even get the system return the value for allo_stat. Thanks in advance for your help!

how to examine filelength?

i'd like to open a file with lines like 1 ; H ; Wasserstoff ; 1.007947 ; 1.0079 2 ; He ; Helium ; 4.0026022 ; 4.0026 i want to find out how many rows this file has. How is that done? Can i use inquire for that? Has someone a good howto for reading data from files? I found a lot of howtos, but they only explained the simple situation without strings and without seperation characters like ";" and the examples expected that the length of the numbers is fixed (like I3). Kind regards, -- Jonas Stein <news@jonasstein.de>

Anyone running Linux and XLF on a PowerMac G5?

Back in ancient times (2005!) I bought a PowerMac G5 running OSX mostly because IBM's XL-Fortran (8.1) was available for it. Shortly afterwards IBM dropped support for it when Apple moved to Intel processors. However, I still use it as my primary development platform becuase it still outperforms the latest Gfortran even though Mac-XLF is getting a little long in the tooth (Linux-XLF is up to version 13.1). I've been thinking of converting the box over to Linux (e.g. YDL 6.2) and getting a later version of XLF. Has anyone gone this route? One stumbling block is that IBM has increas

GAULEG.FOR quadratue weights and nodes.

The routine for finding Gaussian quadrature nodes and weights given in numerical recipe can be trusted for how many digits in double and quad precision.

Calling Linux C library and F2003's C interoperability

Hello, I am trying to figure out if the F2003's C interoperability can help me to call Linux's system library call without resorting to write all sorts of C wrappers for Fortran. Specifically, I want a simple & portable way to create directories from a Fortran program. On Unix / Linux, there is of course a library function that can be called from C: "man 2 mkdir": #include <sys/stat.h> #include <sys/types.h> int mkdir(const char *pathname, mode_t mode); The question is, can I call this from Fortran with the the 2003 C interoperability features to help ? I am not sure w

=?windows-1252?Q?Beginner_Question_=96_passing_strings_to_subroutines?=

Another noob question. I want to be able to pass a fixed string to a subroutine but I can't find any examples on the web and my attempt fails to compile. The program currently looks like this: c-- c This program contains the following examples: c - a simple function that gets user input and returns it. c - a subroutine that prints out the string supplied as a parameter c-- program functions character name*20, get_string*20 name = get_string() call print_string(name) end program functions character*20 function get_string() cha

Returning strings from a function

I am trying to write a simple function that returns a string value. So far it looks like this: c-- c a simple function that gets user input and returns it. c-- program functions character name*20 name = get_string() end program functions character*20 function get_string() character name*20 write (*,'(a)',advance='no') 'Enter your name: ' read *,name get_string = name end I am new to Fortran and am puzzled over what I thought would be a simple task. I get the following error when I try to compile this

Self-referencing abstract interface - legal?

Hello, I happened to make a mistake in defining a set of abstract interfaces and one compiler that I use chokes on it, whereas another accepts it without any fuss. Looking more closely I found the error - the interface "compute_routine" refers to itself and I wanted a different interface. However, I wonder if this is at all legal. Here is the code in question (just the essential part): module computation_data implicit none type computation_method character(len=40) :: name procedure(compute_routine), pointer, nopass :: co

Library files for reading *.HDF files

Hello, everyone! I am working on processing satellite images made in *.HDF format. I have been working with Interactive Data Language (IDL). However, My research is focusing on water resources. Many hydrology and water quality models are developed by FORTRAN, and I think it is meaningful to write up FORTRAN source codes to process images for hydrological application on remote sensing in near future. In order to process satellite images with FORTRAN, I heard I need to have several library files (i.e.:read_hdf.f90 stuff like this). Where can I find this library files. Thanks. Harry

Simply Fortran IDE for gfortran on Windows

http://simplyfortran.com/ Simply Fortran - Affordable Fortran for Microsoft Windows Simply Fortran is a new, complete Fortran solution for Microsoft Windows and compatible operating systems. Designed from the beginning for interoperability with GNU Fortran, Simply Fortran delivers a reliable Fortran compiler on Windows platforms with all the necessary productivity tools that professionals expect. The Simply Fortran package includes a configured GNU Fortran compiler installation, an integrated development environment, a graphical debugger, and a collection of other development nece

call system in MS Fortran 5.1

Hi, I am having troubles regarding "call system" in MS Fortran 5.1. Simply, it doesn't work... I have copied the same code in Microsoft Fortran Power Station 4.0 and everything is working just fine. Linker in MS Fortran 5.1 is saying "error L2029: 'SYSTEM' : unresolved external" . I don't have a clue how to make it work in MS Fortran 5.1. Am I missing some libraries? Can anyone help me? Thank you! p.s. this is the line in my code: call system('c:\print.exe') and print.exe is just printing something on computer screen...

Can't post replies, only new threads

I'm trying to post a reply to the last message in two posts. I have tried 12 times, Normal, short, one line, but nothing works. If this gets through, why? Are there some rules I don't know? Very polite english used. :o)> Can't even report to Google.

click her Russian Girals citizenship and jobs details

Click from Russian Giral Allbamm Beautiful profiles with photos and videos add your proflie for free and start receiving letters and your Jobs details and Passpost details in p.R in Russian citizenship I LOVE ALL RUSSIAN @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ http://123maza.com/25/nike286/ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@

beginner question

Hi, I'm learning Fortran 77, and I have a little problem with I/O. In the following piece of code, the program calls a function u and writes its value on the output. If the "write" is done on a line after the call, I get two lines on output (as expected, I think). But when I write it as shown, then depending on the compiler the program crashes (saying I/O is already open), or prints everything on one line. Is there something wrong in writing code like this ? (e.g., is it forbidden by the standard ?) program test external u integer u,n n=10 wr

how to assign a NaN?

I am reading data with lots of missing values and need a good way to be a ble to assign NaN to integer, real and character vars. the missing value is represented by an empty string '' ! two quotes without a space I am using gfortran which has the isnan(x) to check for nan. I hope there is a way to use a character representation of the NaN so then read(ch(1:2),*) myint,myreal will give me int and real nans from the same string in ch(1:2) thanks everybody.

problem with compiling under G77

Just a short note. I re-wrote the discussed Earthquake data processing program written in Fortan IV for the CDC 6400 in 1978. I finally compiled this with my F77 compiler, linked and ran and it gave the exact same reports, plots and tables as it was expected to. I had kept the intended user abreast of the compile tests (my tentaive version being tested by him because it was initially too large for my compiler to completely compile OR finally link). So after the first crude compile problems were remotely tested with G77 and the error listing information returned to me, I was able