numericaleft / lehmann.jl Goto Github PK

kernel\Omage could return complex or float depending on isFermi and symmetry arguments. Replace the argument with ::Val{isFermi} and ::Val{symmetry} to allow multiple dispatch to improve the type stability

matfreq2dlr, tau2dlr, on axis=3 will transpose other axes.

Hi, I really like this package you have designed! It offers some really cool functionality that is useful in a wide variety of contexts!
However, looking at the docs, many of the doc strings are not formatted correctly and are rendering improperly.
It is very challenging to read the package API as things stand right now.
Would it be possible to have them fixed?

error is Weight, so it can not be 0

Polarization: wrapped: Error During Test at /home/runner/work/ElectronGas.jl/ElectronGas.jl/test/polarization.jl:83
Got exception outside of a @test
InexactError: Int32(1.0e10)
Stacktrace:
[1] Int32
@ ./float.jl:702 [inlined]
[2] (::Lehmann.var"#filename#8"{Symbol})(lambda::Float64, errpower::Int32)
@ Lehmann ~/.julia/packages/Lehmann/uKJiS/src/dlr.jl:102
[3] Lehmann.DLRGrid(Euv::Float64, β::Float64, rtol::Float64, isFermi::Bool, symmetry::Symbol; rebuild::Bool, folder::Nothing, algorithm::Symbol, verbose::Bool)
@ Lehmann ~/.julia/packages/Lehmann/uKJiS/src/dlr.jl:142

When doing fourier transform of objects with dimension D > 1, sumrule needs to be an array with dimension D-1.
When incorrectly using sumrule = 1.0 for an 2D object, the code pops out with:

ERROR: LoadError: MethodError: no method matching reshape(::Float64, ::Int64)
Stacktrace:
[1] tau2dlr(dlrGrid::DLRGrid, green::Matrix{ComplexF64}, τGrid::Vector{Float64}; error::Nothing, axis::Int64, sumrule::Float64, verbose::Bool)
@ Lehmann ~/Julia_manybody3/Lehmann.jl/src/operation.jl:219
[2] tau2matfreq(dlrGrid::DLRGrid, green::Matrix{ComplexF64}, nNewGrid::Vector{Int64}, τGrid::Vector{Float64}; error::Nothing, axis::Int64, sumrule::Float64, verbose::Bool)
@ Lehmann ~/Julia_manybody3/Lehmann.jl/src/operation.jl:390

Dear Kun,

Thank you for the nice Julia implementation of DLR. I am trying to use Lehmann.jl in a project but have stumbled on an issue when calling the DLRGrid builder at high temperatures.

Here is an example code that breaks throwing an ArgumentError

using Lehmann: DLRGrid

for n in range(1, 10)
    β = 1000.0/2^n
    dlr = DLRGrid(Euv=1., β=β, isFermi=true, rtol=1e-9, rebuild=true, verbose=false)
    @show β, n, length(dlr.τ)
end

For the output please see below.

Do you think it would be possible to "harden" the behaviour of the grid builder so that it performs ok also for high temperatures?

Cheers, Hugo

(β, n, length(dlr.τ)) = (500.0, 1, 41)
(β, n, length(dlr.τ)) = (250.0, 2, 37)
(β, n, length(dlr.τ)) = (125.0, 3, 31)
(β, n, length(dlr.τ)) = (62.5, 4, 25)
(β, n, length(dlr.τ)) = (31.25, 5, 20)
(β, n, length(dlr.τ)) = (15.625, 6, 16)
(β, n, length(dlr.τ)) = (7.8125, 7, 13)
ArgumentError: reducing over an empty collection is not allowed

Stacktrace:
  [1] _empty_reduce_error()
    @ Base ./reduce.jl:301
  [2] reduce_empty(op::Function, #unused#::Type{Float64})
    @ Base ./reduce.jl:311
  [3] mapreduce_empty(#unused#::typeof(identity), op::Function, T::Type)
    @ Base ./reduce.jl:345
  [4] reduce_empty(op::Base.MappingRF{typeof(identity), typeof(max)}, #unused#::Type{Float64})
    @ Base ./reduce.jl:331
  [5] reduce_empty_iter
    @ ./reduce.jl:357 [inlined]
  [6] mapreduce_empty_iter(f::Function, op::Function, itr::Vector{Float64}, ItrEltype::Base.HasEltype)
    @ Base ./reduce.jl:353
  [7] _mapreduce(f::typeof(identity), op::typeof(max), #unused#::IndexLinear, A::Vector{Float64})
    @ Base ./reduce.jl:402
  [8] _mapreduce_dim
    @ ./reducedim.jl:330 [inlined]
  [9] #mapreduce#725
    @ ./reducedim.jl:322 [inlined]
 [10] mapreduce
    @ ./reducedim.jl:322 [inlined]
 [11] #_maximum#743
    @ ./reducedim.jl:894 [inlined]
 [12] _maximum
    @ ./reducedim.jl:894 [inlined]
 [13] #_maximum#742
    @ ./reducedim.jl:893 [inlined]
 [14] _maximum
    @ ./reducedim.jl:893 [inlined]
 [15] #maximum#740
    @ ./reducedim.jl:889 [inlined]
 [16] maximum
    @ ./reducedim.jl:889 [inlined]
 [17] testInterpolation(dlrGrid::DLRGrid{Float64, :none}, τ::Lehmann.Discrete.CompositeChebyshevGrid, ω::Lehmann.Discrete.CompositeChebyshevGrid, kernel::Matrix{Float64}, print::Bool)
    @ Lehmann.Discrete ~/.julia/packages/Lehmann/v7X4o/src/discrete/kernel.jl:137
 [18] build(dlrGrid::DLRGrid{Float64, :none}, print::Bool)
    @ Lehmann.Discrete ~/.julia/packages/Lehmann/v7X4o/src/discrete/builder.jl:119
 [19] _build!(dlrGrid::DLRGrid{Float64, :none}, folder::Nothing, filename::String, algorithm::Symbol, verbose::Bool)
    @ Lehmann ~/.julia/packages/Lehmann/v7X4o/src/dlr.jl:338
 [20] DLRGrid(Euv::Float64, β::Float64, rtol::Float64, isFermi::Bool, symmetry::Symbol; rebuild::Bool, folder::Nothing, algorithm::Symbol, verbose::Bool, dtype::Type)
    @ Lehmann ~/.julia/packages/Lehmann/v7X4o/src/dlr.jl:172
 [21] #DLRGrid#7
    @ ~/.julia/packages/Lehmann/v7X4o/src/dlr.jl:189 [inlined]
 [22] top-level scope
    @ ./In[20]:5

so that user don't need to call fourier transform like tau2dlr(:fermi, fdlr, ... ), instead just tau2dlr( fdlr, ...)

In the current implementation dlr2matfreq and tau2matfreq only accept n grid of Int type as the aim grid.
It's more convenient to allow grid of Real number for the purpose of interpolation between different temperatures and perform calculation on an interpolated continuous frequency grid.

_tensor2matrix and _matrix2tensor are type unstable

Which could imply that Lehmann representation failed.

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!