mdav2 / jues.jl Goto Github PK

After merging #32 the following error is being raised:

signal (11): Segmentation fault
in expression starting at REPL[4]:1
in expression starting at REPL[4]in expression starting at REPL[4]:1
jl_apply_generic at /home/aroeira/julia-1.4.1/bin/../lib/libjulia.so.1 (unknown line)
in at ./operators.jl:1042
macro expansion at /home/aroeira/.julia/packages/Lints/II7NR/src/make_ND.jl:26 [inlined]

signal (11): Segmentation fault
in expression starting at REPL[4]:1
in expression starting at REPL[4]:1
ructs.jl:61
#9#threadsfor_fun at ./threadingconstructs.jl:28
_jl_invoke at /buildworker/worker/package_linux64/build/src/gf.c:2144 [inlined]
jl_apply_generic at /buildworker/worker/package_linux64/build/src/gf.c:2322
sig_match_fast at /buildworker/worker/package_linux64/build/src/gf.c:2177 [inlined]
jl_lookup_generic_ at /buildworker/worker/package_linux64/build/src/gf.c:2260 [inlined]
jl_apply_generic at /buildworker/worker/package_linux64/build/src/gf.c:2318
in at ./operators.jl:1042
macro expansion at /home/aroeira/.julia/packages/Lints/II7NR/src/make_ND.jl:26 [inlined]
#9#threadsfor_fun at ./threadingconstructs.jl:61
#9#threadsfor_fun at ./threadingconstructs.jl:28
_jl_invoke at /buildworker/worker/package_linux64/build/src/gf.c:2144 [inlined]
jl_apply_generic at /buildworker/worker/package_linux64/build/src/gf.c:2322
jl_apply at /buildworker/worker/package_linux64/build/src/julia.h:1700 [inlined]
start_task at /buildworker/worker/package_linux64/build/src/task.c:687
unknown function (ip: (nil))
Allocations: 227481754 (Pool: 227388278; Big: 93476); GC: 169
Segmentation fault (core dumped)

Similar issue to #1 . Would be a good time to refactor the CC code anyway.

Trying
pkg> add Combinatorics

Results
Resolving package versions... ERROR: expected package Lints [969b4477] to exist at path /home/aroeira/.julia/packages/Lints/C2dVS

Lints is working fine on my Linux. Previously I had this problem with TBLIS. Is there a issue with git repo on the manifest?

Trying to run JuES directly like:

using JuES
JuES.Input.run("input.dat")

raises an error

ERROR: ArgumentError: NULL library handle
Stacktrace:
[1] #dlsym#1 at /Applications/Julia-1.4.app/Contents/Resources/julia/share/julia/stdlib/v1.4/Libdl/src/Libdl.jl:54 [inlined]
[2] dlsym(::Ptr{Nothing}, ::Symbol) at /Applications/Julia-1.4.app/Contents/Resources/julia/share/julia/stdlib/v1.4/Libdl/src/Libdl.jl:54
[3] TBLIS.TTensor{Float64}(::Array{Float64,4}, ::Float64) at /Users/aroeira/.julia/packages/TBLIS/l7YMy/src/TTensor.jl:51
[4] TBLIS.TTensor{Float64}(::Array{Float64,4}) at /Users/aroeira/.julia/packages/TBLIS/l7YMy/src/TTensor.jl:26
[5] get_eri(::JuES.Wavefunction.Wfn{Float64}, ::String; notation::String, fcn::Int64) at /Users/aroeira/JuES/JuES.jl/src/Backend/IntegralTransformation.jl:74
[6] do_rccsd(::JuES.Wavefunction.Wfn{Float64}; kwargs::Base.Iterators.Pairs{Symbol,Integer,NTuple{5,Symbol},NamedTuple{(:quiet, :do_pT, :return_T, :doprint, :maxit),Tuple{Bool,Bool,Bool,Bool,Int64}}}) at /Users/aroeira/JuES/JuES.jl/src/CoupledCluster/AutoRCCSD.jl:239
[7] exec(::Dict{String,Any}) at /Users/aroeira/JuES/JuES.jl/src/Input/Input.jl:73
[8] run(::String) at /Users/aroeira/JuES/JuES.jl/src/Input/Input.jl:18
[9] top-level scope at REPL[4]:1
[10] eval(::Module, ::Any) at /Applications/Julia-1.4.app/Contents/Resources/julia/lib/julia/sys.dylib:?
[11] eval_user_input(::Any, ::REPL.REPLBackend) at /Applications/Julia-1.4.app/Contents/Resources/julia/lib/julia/sys.dylib:?
[12] run_backend(::REPL.REPLBackend) at /Users/aroeira/.julia/packages/Revise/kqqw8/src/Revise.jl:1163
[13] top-level scope at REPL[1]:0

kwargs passed into correlated methods do not contain options from input.

Example:

Running @show kwargs in the FCI.jl module gives:

kwargs = Base.Iterators.Pairs{Symbol,Integer,NTuple{4,Symbol},NamedTuple{(:quiet, :return_T, :doprint, :maxit),Tuple{Bool,Bool,Bool,Int64}}}(:quiet => true,:return_T => false,:doprint => false,:maxit => 40)

whereas the input is

molecule = """
O        0.6885121727      3.2554927152      0.0182296548
H        1.6778385289      3.2260045029     -0.0128703987
H        0.3991468355      2.6038848620     -0.6689954280
"""
--
options = Dict(
    "reference" => "rhf",
    "basis" => "sto-3g",
    :do_pT => true,
    :frozen => 1,
    :active => 7
)
--
command = JuES.ConfigurationInteraction.FCI.do_fci

After export JULIA_NUM_THREADS

I got the following error as I tried to run a FCI computation

ERROR: TaskFailedException: BoundsError: attempt to access 1-element Array{Any,1} at index [2] Stacktrace: [1] getindex at ./array.jl:788 [inlined] [2] macro expansion at /home/aroeira/.julia/packages/Lints/II7NR/src/make_ND.jl:29 [inlined] [3] (::Lints.var"#9#threadsfor_fun#1"{Array{Float64,4},Array{Any,1},Lints.BasisSetAllocated,UInt64,Array{Any,1},UnitRange{UInt64}})(::Bool) at ./threadingconstructs.jl:61 [4] (::Lints.var"#9#threadsfor_fun#1"{Array{Float64,4},Array{Any,1},Lints.BasisSetAllocated,UInt64,Array{Any,1},UnitRange{UInt64}})() at ./threadingconstructs.jl:28 Stacktrace: [1] wait(::Task) at ./task.jl:267 [2] macro expansion at ./threadingconstructs.jl:69 [inlined] [3] make_ERI(::Array{Float64,4}, ::Array{Any,1}, ::Lints.BasisSetAllocated) at /home/aroeira/.julia/packages/Lints/II7NR/src/make_ND.jl:19 [4] RHFCompute(::JuES.HartreeFock.RHF.RHFWfn; doprint::Bool, maxit::Int64, Etol::Float64, Dtol::Float64) at /home/aroeira/JuES/JuES.jl/src/HartreeFock/RHF.jl:87 [5] RHFCompute at /home/aroeira/JuES/JuES.jl/src/HartreeFock/RHF.jl:77 [inlined] [6] exec(::Dict{String,Any}) at /home/aroeira/JuES/JuES.jl/src/Input/Input.jl:66 [7] run(::String) at /home/aroeira/JuES/JuES.jl/src/Input/Input.jl:20 [8] run() at /home/aroeira/JuES/JuES.jl/src/Input/Input.jl:19 [9] top-level scope at REPL[3]:1 [10] eval(::Module, ::Any) at ./boot.jl:331 [11] eval_user_input(::Any, ::REPL.REPLBackend) at /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.4/REPL/src/REPL.jl:86 [12] run_backend(::REPL.REPLBackend) at /home/aroeira/.julia/packages/Revise/kqqw8/src/Revise.jl:1163 [13] top-level scope at REPL[1]:0

CISingles.jl currently does not correctly compute CIS excitations. Seems to be an issue with indexing the ERIs, possibly a chemists/physicists notation issue. Moreover, the test (CIS energies of H_2 in minimal basis) hid this issue because the ERI tensor is 2x2x2x2 - no bounds violation possible.