Fortran g77 help, 3 questions

[mazinse]

1. Is there a way to get the index of array, how many cells are nonzeros and have actual data. I can probably do it the hard way.

2. How do you pass an array to a function, the entire array I mean, do I do like function (array(i)) or function (array())

3. Is there an easier way to combine two arraries.

If there are libraries or functions out there that I can call to, lease help

 

[jtbell]

1. Is there a way to get the index of array, how many cells are nonzeros and have actual data. I can probably do it the hard way.

Write a do-loop that tests each cell one at a time, and increments a counter each time it finds a nonzero cell.

2. How do you pass an array to a function, the entire array I mean, do I do like function (array(i)) or function (array())

Neither way. Use 'call function (array)'.

3. Is there an easier way to combine two arraries.

Easier than what?

 

[mazinse]

easier than write a loop, checks every cell for value then add them all to a new array.

but I have a new question. When I write my array to a file, I see a lot of 0's, are they empty cells or something else

 

[rcgldr]

With older Fortran compilers, only the addresses of paramters were passed to subroutines. I don't know if this has changed. If it hasn't, then function(array) just passes the address of the array, not the contents of the array.

 

[jtbell]

easier than write a loop, checks every cell for value then add them all to a new array.

I'm pretty sure standard Fortran 77 doesn't have intrinsic functions for array operations (copy, merge, etc.). g77 might add some such functions, but you'd have to search through the g77 documentation.

When I write my array to a file, I see a lot of 0's, are they empty cells or something else

If you stored non-zeroes in all the array cells, then you shouldn't get any zeroes on output. But it's hard to say whether anything is wrong because your description is vague. Can you post a specific code example that demonstrates what you're talking about? It would be best if you can post a short complete compilable program that doesn't do anything else.

 

[TheoMcCloskey]

Also, when you pass an array, you're passing the starting address of the array, ie, a reference to the first element of the array. Hence, for a vector, you can pass any valid "starting address" within the array by specifying the array element, for example

Call MySubProgram(MyVector(2))

or

MyVar = MyFunction(MyVector(2))

This passes a reference to the vector MyVector element 2 to your subprogram or function. Your subprogram or function then sees MyVector starting with element 2 as defined in the main

Also, when passesing multi-dimensioned arrays, you must also pass the physical dimension of the column length so that the subprogram can resolve the array storage layout. For example, passing a two dimensioned array such as

c      MAIN
       REAL  MyArray2(10,20)
c      variable "LDA" is "Leading dimension of Array"
       INTEGER LDA

       ...

       LDA = 10

       CALL MySubProgram(MyArray2, LDA)

       ...

       Subroutine MySubProgram(Array, L)
       INTEGER L
       REAL Array(L,1)
       ...