A simple Julia type for holding Fisher matrices along with parameter names so that they can be added together and the rows and columns are automatically aligned.
To create a Fisher matrix:
julia> f = FisherMatrix([1 0; 0 2],["param1","param2"])
Fisher.FisherMatrix{Int64,ASCIIString}, std-dev of 2 parameters:
param2 => 0.7071067811865476
param1 => 1.0The array itself and the names are stored under f.fish and f.names. We can
create a second fisher matrix and add them together. Note the rows/columns are
automatically aligned.
julia> f2 = FisherMatrix([3 0; 0 4],["param2","param3"])
Fisher.FisherMatrix{Int64,ASCIIString}, std-dev of 2 parameters:
param2 => 0.5773502691896257
param3 => 0.5
julia> f + f2
Fisher.FisherMatrix{Float64,ASCIIString}, std-dev of 3 parameters:
param2 => 0.4472135954999579
param1 => 1.0
param3 => 0.5 The standard deviations (the sqrt's of the diagonal of the inverse) can be
extracted with the stds method
julia> stds(f + f2)
Dict{ASCIIString,Float64} with 3 entries:
"param2" => 0.4472135954999579
"param1" => 1.0
"param3" => 0.5
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent