Negative ESS

[mschauer]

In this example something seems to go wrong (check ESS parameter 4): (data at https://gist.github.com/mschauer/96ffd226d91160b1c9288200c13a77d5)

julia> MCMCChains.Chains(readdlm("badsamples.dat")')
Chains MCMC chain (2000×25×1 reshape(adjoint(::Matrix{Float64}), 2000, 25, 1) with eltype Float64):

Iterations        = 1:1:2000
Number of chains  = 1
Samples per chain = 2000
parameters        = param_1, param_2, param_3, param_4, param_5, param_6, param_7, param_8, param_9, param_10, param_11, param_12, param_13, param_14, param_15, param_16, param_17, param_18, param_19, param_20, param_21, param_22, param_23, param_24, param_25

Summary Statistics
  parameters      mean       std   naive_se      mcse            ess      rhat 
      Symbol   Float64   Float64    Float64   Float64        Float64   Float64 

     param_1    3.4085    1.2065     0.0270    0.0183       955.7380    0.9995
     param_2   -0.5922    0.0681     0.0015    0.0008      4861.8138    1.0003
     param_3    0.0352    0.0087     0.0002    0.0001     10913.2548    0.9995
     param_4   -0.3874    0.0859     0.0019    0.0007   -169293.0177    0.9996
     param_5    0.0045    0.0038     0.0001    0.0001      3396.2860    0.9995
     param_6   -0.2336    0.0598     0.0013    0.0005     23391.6533    0.9997

[mschauer]

Tagging @cscherrer

[mschauer]

Moved to TuringLang/MCMCDiagnosticTools.jl#40