Add support for hybrid basis used in Vlasov simulations #52
Description
In gkylzero we've defaulted polyOrder=1 in Vlasov simulations to mean hybrid basis, meaning p=1 in position and p=2 in velocity space. Currently pgkyl doesn't know how to handle hybrid basis. The problem is that in various places we make decision based on polyOrder or dimensionality, but this has a mixed order, and dimensionality alone doesn't determine the number of basis functions; for example
1x3v
2x2v
both have ndim=4, but they have different numbers of basis monomials. It seems like one would have to make several places aware of cdim and vdim to make this work (or there may be other ways). This is not a tremendous priority because most Vlasov users don't run with p=1/hybrid (although there may be some benefits to doing so in some cases?).
On the other hand there is already support for gkhybrid basis because in that case ndim uniquely specifies the basis (i.e. 1x1v,1x2v,2x2v,3x2v all have a distinct ndim).