Inversion

[prisae]

Inversion capabilities

=> Inversion Milestone
=> invDev

• Analytical gradient
• Jacobian
• A Model-Class that does checks inputs only when necessary (related to bipole without checks & custom example #180)
  • Class can do everything that dipole/bipole/loop/analytical does
  • Checks only done when necessary
  • Logging!
  • ...
• Examples using external tools (pyGIMLi, SimPEG, SciPi, wbi, ...)
• Bring over noise functions from emg3d
• Wrappers to make integration with external tools (pyGIMLi, SimPEG) easy

Attempts in the wild:

• SimPEG: https://github.com/OpenGeophysics/scratchpad
• SimPEG/SciPy: https://github.com/prisae/dev-pygimli-emg3d/blob/tmp/empymod-SciPy-SimPEG.ipynb
• pyGIMLi: https://github.com/gimli-org/transform2021/blob/main/6_Inversion_with_any_forward_operator.ipynb
• pyGIMLi: https://github.com/prisae/dev-pygimli-emg3d/blob/tmp/pyGIMLi-empymod.ipynb
• wbi: https://github.com/WouterDls/1D-wavelet-based-inversion/blob/main/examples/scripts/example_empymod.py
• Work of @mariacarrizo (add links)
• Work of @aignerlukas (add links)

[prisae]

This should be further generalized, so empymod can be used as forward modelling kernel with any inversion tool. An example using pyGIMLi can be found here: https://github.com/gimli-org/transform2021/blob/main/6_Inversion_with_any_forward_operator.ipynb

What is needed: Implement the analytical derivatives.