Hi, when using FixedSizeArrays and doing PCA
using FixedSizeArrays, MultivariateStats
M = FixedSizeArray(randn(10, 100))
fit(PCA, M, method=:svd)
after resolving the errors with MultivariateStats, one is met with
MethodError: no method matching CoefTable(::FixedSizeArray{Float64, 2, Memory{Float64}}, ::Vector{String}, ::Vector{String})
The type `CoefTable` exists, but no method is defined for this combination of argument types when trying to construct it.
Closest candidates are:
CoefTable(::Vector, ::Vector, ::Vector)
@ StatsBase ~/.julia/packages/StatsBase/lRQEN/src/statmodels.jl:72
CoefTable(::Vector, ::Vector, ::Vector, ::Int64)
@ StatsBase ~/.julia/packages/StatsBase/lRQEN/src/statmodels.jl:72
CoefTable(::Vector, ::Vector, ::Vector, ::Int64, ::Int64)
@ StatsBase ~/.julia/packages/StatsBase/lRQEN/src/statmodels.jl:72
...
Stacktrace:
[1] show(io::IOContext{IOBuffer}, ::MIME{Symbol("...")}, M::PCA{Float64})
@ MultivariateStats ~/.julia/packages/MultivariateStats/sSKHh/src/pca.jl:165
Probably CoefTable is hardcoded to expect only a vector. Can this be addressed? Thanks a lot in advance.
Hi, when using FixedSizeArrays and doing PCA
after resolving the errors with MultivariateStats, one is met with
Probably
CoefTableis hardcoded to expect only a vector. Can this be addressed? Thanks a lot in advance.