Rosetta Stone comments

[Beliavsky]

I posted some comments at https://fortran-lang.discourse.group/t/python-fortran-rosetta-stone-ported-from-fortran-90/5907/4

Here are some further comments.

The equivalence of NumPy fancy indexing and Fortran vector subscripts could be discussed. The Python

from numpy import array
a = array([10, 20, 30])
print(a[[0, 2]])

is equivalent to the Fortran

integer :: a(3) 
a = [10, 20, 30]
print*,a([1, 3])

An alternative to the where example in Fortran is a nested merge:

program main
implicit none
integer :: a(10), b(10)
a = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
where (a > 5)
    b = a - 3
elsewhere (a > 2)
    b = 1
elsewhere
    b = 0
end where
print*,b
! lines below added by me
b = merge(a-3, merge(1, 0, a > 2), a > 5)
print*,b
end program main