DensityRatioEstimation.jl in 2D

[MatthewB365]

Hi there, I'm very new to Julia and I'm having some issues with the DensityRatioEstimation.jl package.

I can get things to work for a 1D example however, when I'm trying to extend to 2D I'm running into some issues.

It seems there is a problem with how I'm inputting the data into the function. I have N=1000 samples for each of d=2 parameters for each distribution so I was trying to input the numerator, num, for example as a (1000,2) array - so that each row was a sample and each column was a different parameter. I've also tried the transpose (2,1000) but this didn't work either.

I currently have:

r = densratio(num,den,KLIEP(),optlib=OptimLib)

I'm getting the error:

"MethodError: no method matching *(::Array{Float64, 3}, ::Vector{Float64})
The function * exists, but no method is defined for this combination of argument types."

I've spent a long time looking at the documentation and various sources online but can't seem to find the solution anywhere. I'm currently thinking it may have something to do with the fact that the optimizer only takes vectors as an input?

Any help would be very greatly appreciated!
Thank you.

[juliohm]

Hi @MatthewB365 ,

You need to pass a vector of vectors in num and den. You can achieve this with eachcol(matrix) or eachrow(matrix) for example. These will produce iterators of columns or rows. If it still doesn't work, you can collect these iterators.

I will close the issue, but feel free to ask again if it doesn't work.