Geophysical Inversion and Modeling Library :earth_africa:
panel for 3D pyvista plots in Jupyter Notebook by @andieie in https://github.com/gimli-org/gimli/pull/364
pg.meshtool.createParaMeshPLC3D
pg.show(mesh, "property"),Full Changelog: https://github.com/gimli-org/gimli/compare/v1.2.6...v1.3.0
General
pg.trans.TransSymLog (for signed logarithmically distributed data)Inversion
invert(isReference=True) (by default False! before accidentially True)inv.setRegularization()
correlationLengths=[Ix, Iy, Iz] (so far only global) regularizationModelling
PriorModelling class for both demonstration purposes and incorporation of prior dataLinearModelling based on pg.Modelling for linearized problemsMatrices
RepeatVMatrix, RepeatHMatrix, RepeatDMatrix for vertically/horizontally/diagonally repeated block matricesFrameConstraintMatrix for regularizing (spatially, temporally or spectrally) constraints in frame (multiple models) inversionFrameworks
pg.frameworks.MultiFrameModelling
SIPSpectrum: double Coule-Cole fit with various options (resistivity/Pelton vs. conductivity, tau definition, add vs. mult) ERT: generateDataPDF
Full Changelog: https://github.com/gimli-org/gimli/compare/v1.2.5...v1.2.6
Release presented at EGU2020: https://doi.org/10.5194/egusphere-egu2020-18751
conda install -c gimli -c conda-forge pygimli=1.1.0 # Win, Linux and Mac!
pg.mplviewer - > pg.viewer.mpl
pg.RVector -> pg.Vector
pg.RMatrix -> pg.Matrix
pg.math (cos, cot, sin, det, exp10, median, randn, rms, etc.)pg.matrix (Vector, BlockMatrix, etc.)pg.core.coreMember instead of pg.coreMember
logScale=False
colorBar=True
pg.physics.ert.simulate
pg.physics.ert.createGeometricFactors
pg.fit: run curvefit with ParameterInversion.
Provide function for a curve, data and dataspace and the curve will be fitted with some automagicabs to pg.core objects. e.g. abs(pg.Vector), abs(CVector), abs(R3Vector)
pg.x(obj), pg.y(obj), returns x coordinates for obj (Mesh, DataContainer, R3Vector .. everywhere where we think it can be useful)createPolygon
meshtools.polytools.createCube, createCyinder, extrude, createFacet, createMesh(3dplc)
pg.show with 3D objects (mesh, PLC)pg.Inversion and pg.Modelling are now pure Python.Refraction as TravelTimeManager
ERTManager to replace pybert.Resistivity
pg.unit(), pg.cmap() gives default names and default color maps for your desired physical parameters
only res and vel for now ... will grow on demandpg.debug(), pg.info(), pg.error() and pg.critical(), the latter throws an exception, the other give colored output based on python logger class--debug or -d
@pg.cache
e.g.: @pg.cache
def func(args, **kwargs):
return result
pg.version, and hashable args and kwargspg.core objects or multiple numpy objectspg.core objects--noCache
print(Datacontainer) gives more outputDataContainer and Mesh data values can and should be accessed with index operator
e.g. t = data['t']or mesh['resistivity'] = model
cMap
pg.wait() is obsolete at end of script, any pending MPL figures will be opened
automatically if using a non-interactive backend based on qt or wxconda install -c gimli pygimli=1.0.9 # currently Linux only
Support for secondary nodes in traveltime calculations (following Giroux & Larouche, 2013; http://dx.doi.org/10.1016/j.cageo.2012.12.005)
mesh.createSecondaryNodes(numberOfSecNodes)
rst.setMesh(mesh, numberOfSecNodes)
New method to visualize raypaths: rst.showRayPaths() (https://github.com/gimli-org/gimli/commit/a89eb2c620b3e211e04d63961b4067faedf8b323, API Doc)
pg.show() usage in Jupyter Notebooks) (https://github.com/gimli-org/gimli/commit/207680e86b8a521c735684f673767974202abfc6)cmap and other kwargs in rst.showVA() (https://github.com/gimli-org/gimli/commit/096d02db953f4c1e0224fdefb478c97160de26b2)