Sergey B Kirpichev

Lets add a new kwarg, e.g. ``desc=True`` (or ``reversed=True``) for polyval/polyroots functions to describe the current convention of using *descending* order of coefficients (largest degree - first) with a deprecation...

https://github.com/mpmath/mpmath/pull/779 deprecate descending order of coefficients in polyval() and friends. In v1.5 we can make asc=True - a new default.

> Is the intention that it should be possible to use mpmath functions like det with a numpy array as input? That was my guess as well. > It looks...

> that someone would have a numpy array with dtype=object and mpmath mpfs as the entries in the array? Only this case does make sense for me. But elements of...

> it is a useful feature to allow numpy array to manage a group of mpf objects But this is - a numpy feature, not mpmath. > However, functions in...

> The issue is that the mpmath.matrix class uses a different API convention than numpu array. Really? I don't think so. > ``mp.eigh(a)`` Great! This concrete example shows where is...

> I don't understand why this library is so resistant to numpy. I don't think this is the case. > Leveraging the structure of numpy arrays, or a similar array...

> Creating np.array with mpf type, although displayed as dtype=object, is different to initializing numpy array with dtype=object. You are right, my example should be corrected here. But the conclusion...

> Unless allowing installing numpy in tests It's allowed. Optional dependencies for CI include numpy.

closed in favor of #339