CODEWITHSUNDEEP
fortrangfortranfortran90
user32882
user32882

Reputation: 5897

gfortran default width/decimal number descriptors

The following code demonstrates that for gfortran one may have the processor decide the width of the format descriptor for an integer and floating point number:

program test
    double precision :: x = 5.321568756
    integer ::  i = 53254
    write(*, '(I0)') i
    write(*, '(F0.3)') x
end program test

When compiled, this code outputs

53254
5.322

Is it also possible to have the processor decide on the number of decimal points that are output with the write statement? In other words, to have the above program output the following:

53254
5.321568756

without explicitly providing the number of decimal points in the format descriptor like '(F0.9)'. Something like '(F0.(default number of decimal points here))'

Upvotes: 0

Views: 550

Answers (1)

janneb
janneb

Reputation: 37228

You can use the g0 edit descriptor, like:


program test
  double precision :: x = 5.321568756
  integer ::  i = 53254
  write(*, '(I0)') i
  write(*, '(F0.3)') x
  write(*, '(G0)') x
end program test

(The difference between the G and F edit descriptors is that when the value gets large/small enough, the G edit descriptor switches to exponential notation. Usually that's what you want, though)

Upvotes: 3

Related Questions