Python scripts to plot PECUBE modeling results

Xavier Robert, Grenoble,

This tool have been set up to help the plot of PECUBE (Braun, 2003; Braun et al., 2012) modeling results.

• The Plot-PecubeV4+-Forward.py permits to plot predictions versus observations for Pecube V4+ forward modeling.
• The Inversion_plots-XR.py permits to plot the results of NA Inversions (Sambridge, 1999 a & b) with scatter plots and 1D / 2D marginals. To plot the marginals, the results of NA Inversions should be analyzed with NAB.

Usage

After copying the scripts in your folder ant editing the main sections to fit with your project, each script can be ran in a terminal with the command :

~$ python pyPlotPecubeForward.py

or:

~$ python pyPlotPecubeNA.py

Please, edit the Python files in your text-editor and read the headers of the main section in each files to know how to parametrize them.

The second solution to install the module with pip:

~$ pip install pyPlotPecube

And then, inside a Python environement, import the function you need:

>>> from pyPlotPecube import pyPlotPecubeForward as Fplot

>>> from pyPlotPecube import pyPlotPecubeNA as NAplot

and finally run the function in your Python environement :

>>> Fplot.PlotPecubeForward(dataplot = dataplot, graphpath = graphpath, datafnme = datafnme, inputdata = inputdata, graphtitle = graphtitle, agerange = agerange, profiletype = profiletype, size_x = size_x, size_y = size_y, A = A, B = B)

>>> NAplot.NAplot_Pecube(param = param, dataplot = dataplot, inv_results = inv_results, data_nab = data_nab, graph_path  = graph_path, PDF_1D = PDF_1D, PDF_2D = PDF_2D, pdf1d_results = pdf1d_results, pdf2d_results = pdf2d_results, tick_space = tick_space, size_x = size_x, size_y = size_y, size_m = size_m, size_mis = size_mis)

You can also write a script that facilitate the call of the function with the right input data. See Templates in Tests/.

Parameters

PlotPecubeForward

For Forward plots, arguments are :

1. datafnme (string): results of Pecube forward modeling.

   No default value.

   2. inputdata (string): input data declared in Pecube.in. This is used to plot the errorbars.

   No default value.

3. inputdataPTt (str, optional): Path and file name of the input PTt data file. Need to be given if 'TTp' in dataplot.

   Default = None .

4. outputdataPTt (str, optional): Path and file name of the output PTt prediction/comparison file. Usually, this is 'output/CompareTT.csv'. Need to be given if 'TTp' in dataplot.

   Default = None.

5. stackTT (boolean, optional): Tells if we want to plot all the PTt path in one single graph

   Defaults = True.

6. inputdataFTL (str, optional): Path and file name of the input FTL data file. Need to be given if 'FTL' in dataplot.

   Default = None.

7. outputdataFTL (str, optional): Path and file name of the output FTL prediction/comparison file. Usually, this is 'output/CompareFTL.csv' Need to be given if 'FTL' in dataplot.

   Default = None.

   8. dataplot (list, optional): List of data to plot: ['AHe', 'AFT', 'ZHe', 'ZFT', 'KAr', 'MAr', 'BAr', 'MTL', 'TTp']; by default, the altitude will be plotted; Do not forget the simple quotes !!!

   Note: For the moment, MTL is not implemented.

   If 'TTp', you should give the PTT files names (inputdataPTt and outputdataPTt)

   Defaults = ['AHe','AFT'].

   9. graphpath (str, optional): name of the folder where the plot will be written. Usually you do not have to change it.

   Defaults = 'Graphs'.

   10. graphtitle (str, optional): title to write on the graph.

   Defaults = None.

   11. agerange (2*1 array of floats, optional): range of the ages to plot on the profiles [min, max].

   Defaults = None.

   12. profiletype (list, optional): type of profile, could be one or more of ['Latitude', 'Longitude', 'Altitude', 'Projected']. If [], no age profile is plotted.

   Defaults = [].

13. size_x (int, optional): Font size for x-axis.

    Defaults = 15.

    14. size_y (int, optional): Font size for y-axis.

    Defaults = 15.

    15. A, B (floats, optional): If need of a projected transect, define the line along which we will project with the coordinate of the point A and B defining, respectivelly, the begining and the end of the transect, in lat-long/WGS84.

    Defaults = None.

16. agename (dict, optional): legend of each data system

    If None, this is set to

    {'AHe' : 'AHe (Ma)',

    'AFT' : 'AFT (Ma)',

    'ZHe' : 'ZHe (Ma)',

    'ZFT' : 'ZFT (Ma)',

    'KAr' : 'KAr (Ma)',

    'BAr' : 'Biot. Ar (Ma)',

    'MAr' : 'Musc. Ar (Ma)',

    'HbAr' : 'Hb Ar (Ma)',

    'FTL' : 'FT length (µm)'}

    Default = None.

17. predname (dict, optional): legend of each predicted system.

    If None, this is set to:

    {'AHe' : 'Predicted AHe (Ma)',

    'AFT' : 'Predicted AFT (Ma)',

    'ZHe' : 'Predicted ZHe (Ma)',

    'ZFT' : 'Predicted ZFT (Ma)',

    'KAr' : 'Predicted KAr (Ma)',

    'BAr' : 'Predicted Biot. Ar (Ma)',

    'MAr' : 'Predicted Musc. Ar (Ma)',

    'HbAr' : 'Predicted Hb Ar (Ma)',

    'FTL' : 'Predicted FT length (µm)'}

    Default = None.

18. colores (dict, optional): Colors used for the different age system

    If None, this is set to:

    {'AHe' : 'y',

    'AFT' : 'r',

    'ZHe' : 'g',

    'ZFT' : 'b',

    'KAr' : 'k',

    'BAr' : 'c',

    'MAr' : 'm',

    'HbAr' : '0.75',

    'FTL' : 'y'}

    Default = None.

NAplot_Pecube

For Inverse plots, arguments are :

1. param (list of str): Define as many variable as you have, with their unit. Check the order in na.sum (open it with a text editor) or in NA_Results. If you use the later, first column is always the misfit. /!If you want to use a slash`, use this one --> ``'∕' <-- . The regular slash it is a unicode division symbol windows and OSX do not allow the use of it.

   Exemple : param = ['Offset (km)','Basal Temperature (°C)','Slip rate (km∕Ma)']

2. dataplot (list of couple of integers): Set the couple of variables to plot against each other. If you plot 2D pdfs (contours), please, CHECK that the couple of parameters to plot are the same and in the same order than in the nab.in file. This python script checks it and will insult you if this is not compatible !!!

   Exemple : Offset vs Slip rates, dataplot = [(1,3)] --> plot=(1,3);

   No default value.

3. tick_space (array of floats): Set the space between ticks for x and y axes for each parameters (same order than the list param). If the tick format does not fit your variables, you may need to modify the dictionnary tick_order in the function multiplot.

   No default value.

4. inv_results (str, optional): Name of the NA file with the inversion results, usually NA_Results.

   Defaults = 'NA/NA_results.csv'.

5. data_nab (str, optional): Name of the NAB file with the inversion results, usually nab.out.

   Defaults = 'NA/NAB/nab.out'.

6. graph_path (str, optional) : Path where to save graphs and results Usually NA/Graphs.

   Do not forget the '/' at the end.

   Defaults = 'NA/Graphs/'.

7. PDF_1D (bool, optional): Choose if you want the 1-PDFs (Probability Density Function);

   Defaults to True.

8. PDF_2D (bool, optional): Choose if you want the 2-PDFs (Probability Density Function);

   Defaults = False.

9. pdf1d_results (string, optional): Print the 1-pdfs in a text file.

   Defaults = None.

10. pdf2d_results (string, optional): Print the 2-pdfs in a text file.

    Defaults = None.

11. size_x (int, optional): Size of the font for the x axes label.

    Defaults = 15.

12. size_y (int, optional): Size of the font for the y axes label.

    Defaults = 15.

12. size_m (int, optional): Size of the font for the markers label.

    Defaults = 15.

13. size_mis (int, optional): Size of the markers of the misfits.

    Defaults = 50.

14. peak_thres (float, optional): Threshold to find peaks; between 0. and 1. See peakutils documentation.

    Default = 0.05.

15. peak_min_dist (interger, optional): Minimum distance between the peaks. See peakutils documentation.

    Default = 30.

Examples

Two example's sripts are in the Tests/ folder. Just run them from their location :

~$ python Template_ForwardPlot.py

~$ python Template_NAPlot.py

For instance, the Template_ForwardPlot.py script gives the plot:

Figure 1: Data and predictions along a projected transect.

and the Template_NAPlot.py script permits to build the plots:

Figure 2: NA inversion results for 2 parameters; on the scatter plot, each point corresponds to a model, and the color corresponds to the value of the misfit for that model; The curves on the sides shows the 1D-pdf of each parameter.

Figure 3: Curve fitting of the 1D-pdf of one parameter after NA inversion.

How to cite

Please, if you use this module, cite : Robert X., pyPlotPecube, a python module to plot PECUBE forward and inverse modeling results (2021), DOI:10.5281/zenodo.5521061

Contact

If needed, do not hesitate to add a new branch or to contact the author. Please, use https://www.isterre.fr/identite_id135055.html#

Licence

Copyright (c) 2021 Xavier Robert

This package is licenced with <SPDX-License-Identifier: GPL-3.0-or-later>