Double precision array error

Question

I'm trying to run code for a mathematical algorithm (Conjugate Gradient method). In doing so I input a double precision matrix, defined as such in my preamble. When compiling, I get the follow error:

A=RESHAPE((/ 0,8,0,4,26,8,0,17.5,0,0,0,17.5,0,2.5,-8,4,0,2.5,0,-5,26,0,-8,-5,0 
                        1
Error: Element in INTEGER(4) array constructor at (1) is REAL(4)
make: FTranProjectBuilder: Error: Execution exited with code 2 
*** [cg_main.o] Error 1

My definition in the program with the matrix being defined is given as such (the array definition is the first operation of my program):

PROGRAM cg_main 
IMPLICIT NONE 

INTEGER,PARAMETER                     ::d=5 !use a parameter for the dimensions (simple)
DOUBLE PRECISION,DIMENSION(d,d)       ::A !matrix
INTEGER,DIMENSION(2)                  ::order2 = (/ 2, 1 /) !matrix reshape order

[MORE DECLARATIONS HERE]

A=RESHAPE((/ 0,8,0,4,26,8,0,17.5,0,0,0,17.5,0,2.5,-8,4,0,2.5,0,-5,26,0,-8,-5,0 /),(/d,d/), order2) !specify dxd matrix

[MORE CODE HERE]

END PROGRAM

The code works without the decimal numbers in my matrix input, but doesn't seem to with my decimals and I have no idea why.

francescalus · Accepted Answer · 2022-11-12 21:02:01Z

2

All elements in an array constructor of the form of the question must be of the same declared type and same kind type parameters. The compiler here complains because this constraint has been violated by mixing a real expression in with those integer expressions.

Simply making the first element a real/double precision element is not sufficient. Perhaps the compiler is using the error message about an "integer(4) array constructor" as short hand for "the first element is integer(4) but not all other elements are".

To resolve this, with this form of the array constructor you should make all elements of the same type/kind (probably real/double precision).

There is also another form for array constructors:

[ real ::0,8,0,4,26,8,0,17.5,0,0,0,17.5 ... ]

which converts all elements to the type specified. You can even use this form to create a size-zero array:

 [ real :: ]

edited Nov 12, 2022 at 21:02

answered Oct 5, 2017 at 8:26

francescalus

32.9k28 gold badges86 silver badges128 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Double precision array error

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related