Progress in refactoring code and implementing unit tests #82
Conversation
src/m_common/upsim.f90
Outdated
| logical symmagp | ||
| ! argument parameters | ||
| integer, intent(in) :: ndim, ind, ipf | ||
| real(8), intent(in) :: value |
There was a problem hiding this comment.
real(8) and real*8 is not the same. Well it is in this case but not in the complex case. Its good to stick with one form. In most of the core star convetion is use.
There was a problem hiding this comment.
Hi @addman2 Thank you for the suggestion! I used the same conventions as the old ones. I agree to stick with one form! Shall we use the core star convention (real*8, complex*16 etc...) in all files? I can do this change while refactoring the code.
There was a problem hiding this comment.
Hi @kousuke-nakano ,
Yes, I think the star convention is already used more than the bracket convention. So to change to star one is easier than other way around. I can add this to the linter.
There was a problem hiding this comment.
Dear @addman2, sure! I will use the star convention in all sources. I will fix it during the ongoing refactoring process.
There was a problem hiding this comment.
Dear @kousuke-nakano,
what is the problem with intel OneAPI compilation? I know there are issues with it which one you experience?
There was a problem hiding this comment.
HI @addman2, the problem was the Intel GPG key. I guess it was updated. The problem has been solved. see #105
src/m_common/upvpot_ei.f90
Outdated
|
|
||
| ! argument parameters | ||
| integer nion | ||
| real(8), intent(in) :: rkel(3) |
There was a problem hiding this comment.
Same here, see comentary of upsim.f90
There was a problem hiding this comment.
Same above.
src/m_common/upvpotaa.f90
Outdated
|
|
||
| ! argument parameters | ||
| integer, intent(in) :: nion | ||
| real(8), intent(in) :: zeta(nion), iond(nion, nion), kappa, LBox |
|
Hi all,
I think the bracket convention is more used, or at least is the one I have
seen used more often in the most recent fortran codes.
Ciao
Michele
…On Sat, Sep 30, 2023 at 4:29 PM kousuke-nakano ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In src/m_common/upsim.f90
<#82 (comment)>:
> @@ -17,9 +17,15 @@ subroutine upsim(amat, ndim, ind, value, symmagp, ipf)
use allio, only: nelorb_at, pfaffup, kiontot
implicit none
- integer ndim, ndimh, ind, i, j, ipf
- real(8) amat(ndim, *), value
- logical symmagp
+ ! argument parameters
+ integer, intent(in) :: ndim, ind, ipf
+ real(8), intent(in) :: value
Dear @addman2 <https://github.com/addman2>, sure! I will use the star
convention in all sources. I will fix it during the ongoing refactoring
process.
—
Reply to this email directly, view it on GitHub
<#82 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AALBETJC3LTZE2MLJNYKK3LX5AUGDANCNFSM6AAAAAA5HJPHRM>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
|
Dear @michele-casula, yes, you are right. The bracket one seems newer. Dear @addman2, we can use the bracket one, what do you think? As you pointed out, we should be careful about the complex case, i.e., complex(8) = complex*16. |
What I am doing right now are:
In this pull request, I have changed several codes in /src/m_common/
#12