john verzani
john verzani
> This is a short account of my first experince moving to the new `Polynomials`. > > I started to modify the software in [MatrixPencils.jl ](https://github.com/andreasvarga/MatrixPencils.jl) to comply with the...
> With the new structure, we could consider the following structures: > > ``` > > julia> tx=AbstractVecOrMat{Polynomial{T,X} where T} where {X} > Union{AbstractArray{Polynomial{T,X} where T,1}, AbstractArray{Polynomial{T,X} where T,2}} where...
Yes indeed. The issue is promote between containers with polynomials. In this example there is no promotion between Tuple{P{Int,:X},2) and Tuple{P{Float64,:X},2) which is understandable, tuples and all. But I naively...
I'll have to look. My guess would be that if you do one of three things some workaround will be needed: * subtype `AbstractPolynomial{T}` (in which case `AbstractPolynomial{T,X}` is needed...
Should this be a special method for `eigvals` in base? If that isn't appropriate, I'd be interested in the PR here or in `PolynomialZeros`.
Here is a *painful* way to use this class: ``` X = sympy"MatrixSymbol"("X",3,3) Y = sympy"MatrixSymbol"("Y",3,3) ``` So far so good. But to use them is cumbersome: ``` PyCall.py"$(X[:T]()) *...
Well, I tried, but didn't get this done. Following the above comment leads to a situation where simple things work fine, but ultimately fail as the two types (SymMatrixSymbol and...
Auto-converting one-dimensional AbstractArray to Matrix breaks type-checking logic in Python library
Yes, reasonable point caused by https://github.com/JuliaPy/SymPy.jl/blob/master/src/SymPy.jl#L128 Do you know if vectors are a distinct python class?
Auto-converting one-dimensional AbstractArray to Matrix breaks type-checking logic in Python library
I think this happens in PyCall (the conversion from PyList to an array) in the conversions.jl file. I'm not sure how to get a List type to flow through. What...
Thanks for the report! A fix is coming soon.