notice that you set x(i+1) but use x(i) in the next line; you need a better way to connect the values at each step: n, n-1, n-2, ..., 3,2,1. You can loop over values in decreasing order; if you wanted to be much more careful you might note that n! is integer and set "fact" as an integer

fact = 1

do i=n1,1,-1                       [the -1 is the step]

fact = fact*i

end do

there is no need to store the value that you get at each iteration in a vector x

I know fortran but not touched since long time.  But I can suggest it. I had worked Fortran in WATFOR compiler long back. Once suggestion is.

Declare the statement as DIM A(5). Some languages dont take the dimesion of the array through the keyboard . The dimension must be specified eartlier. 

In C also we cannot use  int a[r][c] but we have to provide the values directly atleast by Preprocessor directive. SO . But in fortan there is no preprocessor directive. So check the grammer of arrays in FORTRAN.

Hi,

Some suggestions:

1. Make the array of a constant, but bigger size, say 20. As suggested, most languages do not accept variable size arrays.

2. No need for arrays here at all. Try multiplying 1, 2, ..., n without storing them in an array, if your goal is just writing a factorial program (and not learning FORTRAN).

notice that you set x(i+1) but use x(i) in the next line; you need a better way to connect the values at each step: n, n-1, n-2, ..., 3,2,1. You can loop over values in decreasing order; if you wanted to be much more careful you might note that n! is integer and set "fact" as an integer

fact = 1

do i=n1,1,-1                       [the -1 is the step]

fact = fact*i

end do

there is no need to store the value that you get at each iteration in a vector x

Thank You Morton! That worked! I learnt a new thing regarding do loop today!

Hi Mousumi,

This isn't the best way to compute a factorial, but that has nothing to do with the issues.

When you declare an array, the compiler has to allocate memory for it.  You have to specify constant values for the bounds.  The only exception is when the array is allocated at run-time.  Then you declare the array to be allocatable and you specify the shape but not the bounds.  You set the array up with an ALLOCATE statement - see the attached file.

By the way, to prompt for input you would like to have non-advancing I/O.  The standard way to do this is cumbersome and I have never seen it used in anger.  Best to use the non-standard, but universally supported $ format descriptor - see the attached file.

Best wishes,

John

"In C also we cannot use int a[r][c] but we have to provide the values directly atleast by Preprocessor directive."

This is not correct. The C language now supports VLA -- variable length arrays.

As for FORTRAN, John Collins appears to have the correct answer.

I am satisfied with the answers by John Collins and Jim Balter

This may help you :

https://stackoverflow.com/questions/45645815/how-to-implement-factorial-function-into-code