On 2/10/21 2:39 AM, Thomas de Groot wrote:
> Op 09/02/2021 om 13:56 schreef William F Pokorny:
>> On 2/8/21 2:38 PM, Bald Eagle wrote:
...
> [snip]
> 
> Wow! Some nice piece of info there!
> 

Keep the disclaimer(a) in my post in mind. I believe most of the 
information is good, but should I get back to digging around again in 
that code, some of it will need revision.

(a) - I'm talking too about the current v3.8. Christoph made changes 
relative to v3.7.

> I was late testing this media and found also that samples were the 
> culprits.
> 
> I have one question about them: can samples still be used with two 
> parameters with method 3? The Wiki is confusing about this, and I 
> wondered if it was a legacy from method 1 or 2, but not necessary anymore.
> 

Yes. You can have the max value, it doesn't hurt anything(b). It looks 
to be used in the statistical sampling methods.

---
(b) Your question reminded me of another media issue. We can have 
multiple medias. When we do there were situations I cannot quite recall 
- with ordering... Mixing media 1&2 with 3 - 'maybe.'

In some of those cases, depending on ordering, certain settings matter 
which would not if a particular media in the collection were run alone. 
It's been too long - I'm not remembering the details, but there were 
issues with multiple medias and option settings.

Hmm, another ringing bell; Was there also something with one media in 
multiples being constant and others not...?

I might have created test cases for some of the issue/questions. I have 
many media test cases, though, most were for work on the solvers. Media 
in containers make good solver test cases because you have to get all 
the roots right...

Anyway, I'm focused elsewhere at the moment.

Bill P.

Post a reply to this message