MethodError using Omniscape

General Usage
1

Hi all,

I’m working on running an analysis in Omniscape, and am battling an error that I will post below. I have run these analyses before without issue. I’m getting this issue even when trying files that worked successfully a few months ago, and on different computers. The resistance, source layers, and .ini file looks fine.

Any help is very much appreciated!

julia> run_omniscape(“chase_config.ini”)
┌ Warning: Your specified project directory, outputs/chase_local_movement, already exists. Writing outputs to outputs/chase_local_movement_2.
└ @ Omniscape C:\Users\peteb.julia\packages\Omniscape\VSLPS\src\main.jl:155
[ Info: Starting up Omniscape with 8 workers and double precision
[ Info: Using Circuitscape with the cg+amg solver…
[ Info: Calculating block artifact correction array…
ERROR: MethodError: no method matching missingarray_to_array(::Matrix{Union{Missing, Float64}}, ::Int64)
This error has been manually thrown, explicitly, so the method may exist but be intentionally marked as unimplemented.

Closest candidates are:
missingarray_to_array(::Array{Union{Missing, T}, N} where {N, T<:Number}, ::Number)
@ Omniscape C:\Users\peteb.julia\packages\Omniscape\VSLPS\src\utils.jl:614

Stacktrace:
[1] calc_correction(arguments::Dict{…}, os_flags::Omniscape.OmniscapeFlags, cs_cfg::Dict{…}, condition_layers::Omniscape.ConditionLayers{…}, conditions::Omniscape.Conditions, precision::DataType)
@ Omniscape C:\Users\peteb.julia\packages\Omniscape\VSLPS\src\utils.jl:471
[2] run_omniscape(cfg::Dict{…}, resistance::Matrix{…}; reclass_table::Nothing, source_strength::Matrix{…}, condition1::Nothing, condition2::Nothing, condition1_future::Nothing, condition2_future::Nothing, wkt::String, geotransform::Vector{…}, write_outputs::Bool)
@ Omniscape C:\Users\peteb.julia\packages\Omniscape\VSLPS\src\main.jl:252
[3] run_omniscape(path::String)
@ Omniscape C:\Users\peteb.julia\packages\Omniscape\VSLPS\src\main.jl:558
[4] top-level scope
@ REPL[12]:1
Some type information was truncated. Use show(err) to see complete types.

2

Odd, that call signature looks like it would dispatch to that method, can you check that file and line for what the method body is?

3

Hi Benny,

Sorry I don’t totally know what you mean by that. But here is the .ini file that I’m getting the error on (this just worked perfectly a month ago) if this helps.

[Input files]
resistance_file=resistance_map.tif

[Options]
project_name=outputs/chase_local_movement
block_size=13
radius=313
source_from_resistance=true
write_raw_currmap=true
calc_normalized_current=true
calc_flow_potential=true
mask_nodata=true
write_as_tif=true
parallelize=true

4

Ah never mind, here is the code from that file at line 614 onwards

*nodata**: The numeric value with which missing values will be replaced in
the result.
“”"
function missingarray_to_array(
A::MissingArray{T, N} where T <: Number where N,
nodata::Number
)
output = copy(A)
output[ismissing.(output)] .= nodata

return convert(Array{typeof(output[1]), ndims(output)}, output)
5

By all appearances of the stacktrace, that method dispatch should have worked, and that method works on simple inputs when copied and isolated to a different module. There’s an open issue for this, so if you have a Github account you could chime in there as another user. That issue’s commenters say that downgrading to Julia 1.11 works in the meantime, though I can’t imagine how 1.12 made a difference.

method dispatch error when running forest connectivity example · Issue #160 · Circuitscape/Omniscape.jl

Just curious, does this run properly for you on 1.12?

module Blah
const MissingArray{T, N} = Array{Union{Missing, T}, N}
function missingarray_to_array(
        A::MissingArray{T, N} where T <: Number where N,
        nodata::Number
    )
    output = copy(A)
    output[ismissing.(output)] .= nodata

    return convert(Array{typeof(output[1]), ndims(output)}, output)
end
end

Blah.missingarray_to_array([0.1 missing                                                                                                                                                                
                            2.3 missing], -9999)

import Omniscape

Omniscape.missingarray_to_array([0.1 missing                                                                                                                                                                
                                 2.3 missing], -9999)