Edit on GitHub

SlimPlotting

Utility function to plot physically gridded (seismic) data in the time and frequency domain and model parameters and perturbations. The functions currently only supports 2D data.

Disclaimer

This is a small package that I started to make my life easier visualizing data and results easily. Not all functionalities are tested and some of the features may need additional options for better images. This package is under development and welcomes contributions through Issues, Pull Requests or Discussions

Functionalities

This package implement four main functions that rely on a base _plot_with_units internal function.

  • plot_simage to plot a 2D seismic image (i.e RTM)
  • plot_sdata to plot 2D seismic data such as a shot record
  • plot_fslice to plot a 2D frequency slice of seismic data.
  • plot_velocity to plot a 2D velocity model.
  • wiggle_plot to make a 2D wiggle plot of a seismic data.

The functions plot_simage, plot_sdata and plot_velocity support abstract object with meta-data containing the grid spacing. For example, you can plot a 2D JUDI PhysicalParameter either via plot_velocity(p.data, p.d) or directly via plot_velocity(p) that will extract the data and spacing automatically. While this supports JUDI since we are using it extensively, this package does not depend on it and only expect a Julia structure as an input (when the spacing is not specified) containing a .d attribute with the grid spacing. We show in the simple example how to setup such a simple structure.

The expected inputs are:

  • plot_simage(array, tuple; kw...) or plot_simage(structure) with strucutre.d

containing the grid spacing and structure.data containing the 2D array.

  • plot_velocity(array, tuple; kw...) or plot_velocity(structure) with strucutre.d

containing the grid spacing and structure.data containing the 2D array.

  • plot_sdata(array, tuple; kw...) or plot_sdata(structure) with strucutre.dt

containing the time sampling rate and structure.d containing the receiver spacing (uniform sampling is assumed at the time) and structure.data containing the 2D array.

  • wiggle_plot(array, xrec, time_axis; kw...). In this case, xrec, time_axis are optional and wil default to 1:size(array, 2), 1:size(array, 1) respectively.

and you can check the docstring (julia ?) for additional information on the optional keyword arguments.

Color maps

The colormap support is extensive as this uses three sources:

  • matplotlib standard colormaps.
  • ColorShemes.jl that implements a variery of colormaps from different packages including Matplotlib, Seaborn, GNUPlot, colorcet(Collection of perceptually accurate colormaps). You can provide the chosen colormap as a kewyword , i.e plot_simage(array, tuple; cmap=:jet).
  • colorcet perceptually accurate colormaps that are available through their colorcet names (i.e cet_rainbow4 for a perceptually accurate jet colormap).
  • seismc a set of seismic colormap usable via seiscm(name) where name is one of the four supported maps seismic, bwr, frequency, phase

All functionality, with the exeption of wiggle_plot accept the keyword argument cmap, i.e plot_velocity(array, spacing; cmap=:vik).

Authors

This package is developed and maintained by Mathias Louboutin<mlouboutin3@gatech.edu> and the ML4Seismic Lab at Georgia Institute of Technology.