Python 3 compatibility with dictionary views vs lists.

[rjkoch]

So, if I try to run a co-refinement, where I have more than one fit contribution, generally the recipe building goes fine.

But, I if call something like FitResults(recipe).printResults() I get:

...
AttributeError: 'odict_keys' object has no attribute 'sort'

Something similar occurs if I call recipe.setWeights

I think this is because dictionary keys/values etc are being treated as lists (python 2)

I'll see if I can fix it.

[sbillinge]

Thanks Rob. It looks as if it might be an ordered dict, is it? You can't sort a dict, but you may be able to sort an ordered dict. I think those are in package containers. Anyway, look into it, we may be able to get the same behavior this way, though I am just guessing.

…

On Mon, Oct 28, 2019 at 10:37 AM Robert J Koch ***@***.***> wrote: So, if I try to run a co-refinement, where I have more than one fit contribution, generally the recipe building goes fine. But, I if call something like FitResults(recipe).printResults() I get: ... AttributeError: 'odict_keys' object has no attribute 'sort' Something similar occurs if I call recipe.setWeights I think this is because dictionary keys/values etc are being treated as lists (python 2) I'll see if I can fix it. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#60?email_source=notifications&email_token=ABAOWUIHFU3BD6GO7U6HDJ3QQ32KVA5CNFSM4JF3C3Y2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HUZBB2Q>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABAOWUJOD47QU5M4DV5T7ALQQ32KVANCNFSM4JF3C3YQ> .

[rjkoch]

The issue is that '.sort()' and '.index()' are being called on the return of 'dict.keys()' and 'dict.values()' In python 2, 'keys()' and 'values()' return lists, so '.sort()' and '.index()' are fine. In python 3, 'keys()' and 'values()' do not return lists, so this breaks. The simplest solution I think is to wrap these 'dict.keys()' and 'dict.values()' calls in 'list()' I've already fixed the two cases I encountered in a cloned local repo. I just need to figure out how to submit a pull request (it's will be my very first). The more difficult part will be implementing this in method/function returns that are then used by other functions. It's not clear to me when the return should be a list, a view, or a generator. Rob On Mon, Oct 28, 2019 at 10:40 AM Simon Billinge <notifications@github.com> wrote:

…

Thanks Rob. It looks as if it might be an ordered dict, is it? You can't sort a dict, but you may be able to sort an ordered dict. I think those are in package containers. Anyway, look into it, we may be able to get the same behavior this way, though I am just guessing. On Mon, Oct 28, 2019 at 10:37 AM Robert J Koch ***@***.***> wrote: > So, if I try to run a co-refinement, where I have more than one fit > contribution, generally the recipe building goes fine. > > But, I if call something like FitResults(recipe).printResults() I get: > > ... > AttributeError: 'odict_keys' object has no attribute 'sort' > > > Something similar occurs if I call recipe.setWeights > > I think this is because dictionary keys/values etc are being treated as > lists (python 2) > > I'll see if I can fix it. > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > < #60?email_source=notifications&email_token=ABAOWUIHFU3BD6GO7U6HDJ3QQ32KVA5CNFSM4JF3C3Y2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HUZBB2Q >, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/ABAOWUJOD47QU5M4DV5T7ALQQ32KVANCNFSM4JF3C3YQ > > . > — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#60?email_source=notifications&email_token=AMSFZJAIMIOH7SFNDL5GBHDQQ32WBA5CNFSM4JF3C3Y2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECNDLWI#issuecomment-546977241>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AMSFZJBIDNP4LEFXB352RRDQQ32WBANCNFSM4JF3C3YQ> .

[rjkoch]

Awesome! Thanks! Rob

…

On Thu, Nov 14, 2019 at 3:15 AM Pavol Juhas ***@***.***> wrote: Closed #60 <#60> via #61 <#61>. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#60?email_source=notifications&email_token=AMSFZJEVNMIOWZ65MKJMMILQTUCK5A5CNFSM4JF3C3Y2YY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOU3D22GA#event-2798103832>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AMSFZJAUHWNSRT7DHJYBFELQTUCK5ANCNFSM4JF3C3YQ> .