1d input of a LayeredModel from EvalModel

[Thomas-Ulrich]

Hi,
I would like to define the 1d input of a LayeredModel with an EvalModel combining 2 inputs (in the real-case application, one input will be from an ASAGI component). This does not work. Do you think is it a limitation of the component or a bug?

Thanks,

Thomas.

test.yaml

!Switch
[plastCo]: !LayeredModel
      map: !Include test2.yaml
      interpolation: linear
      parameters: [plastCo]
      nodes:
        10e3:  [222339.89064034278]
        -10e3: [354563.38402222196]

test2.yaml

!EvalModel
parameters: [elevation,z]
model: !Switch
  [elevation]: !ConstantMap
    map: 
        elevation: 10.
  [z]: !AffineMap
    matrix:
      z: [0.0, 0.0, 1.0]
    translation:
      z: 0.0
components: !FunctionMap
    map:
      z_below_surface: |
        return elevation + z;

Error message:

di73yeq4@login03:/hppfs/work/pr63qo/di73yeq4/SulawesiSetupNewYann> easicube '{from: [0,0,150], to: [1,1,250], N: [1,1,10], group: 1}' test.yaml 
test.yaml: yaml-cpp: error at line 2, column 12: Layered model requires 1D input (got elevation,z).
!<!LayeredModel>
map: !<!Include> test2.yaml
interpolation: linear
parameters: [plastCo]
nodes:
  10e3: [222339.89064034278]
  -10e3: [354563.38402222196]
terminate called after throwing an instance of 'std::runtime_error'
  what():  Error while parsing easi model file.
Aborted (core dumped)

[krenzland]

Hm, I'm not too familiar with the layered model but it sounds like it is confused about which input it should be interpolating - there's both elevation and z but you can only use this model with one coefficient.
That's at least what the documentation (and the error message) is saying:
https://easyinit.readthedocs.io/en/latest/builders.html?highlight=LayeredModel#layeredmodel

[Thomas-Ulrich]

Hum 'root component', I see ...

[Thomas-Ulrich]

I guess I can use a 1D ASAGI instead of the LayeredModel

[krenzland]

Maybe you could also interpolate using a function/Lua map? For two nodes this shouldn't need too much code.

[Thomas-Ulrich]

I have 150 nodes...