doi-usgs / geobipy Goto Github PK
View Code? Open in Web Editor NEWGeophysical Bayesian Inference in Python. Docs:
Home Page: https://usgs.github.io/geobipy/
License: Other
Geophysical Bayesian Inference in Python. Docs:
Home Page: https://usgs.github.io/geobipy/
License: Other
it would be great if gaps in line plots were shown as actual gaps rather than extrapolating the closest models (on either side of the gap) into the gap.
The hyperbolic tangent in Fortran is not defined in the same way on different systems. On my Mac, high positive values produce 1+0j, while on sci linux I got Nan+0j. Same version of gfortran.
I think the datasets (.txt) files have disappeared from the supplementary/Data folder
For users running this code on windows, there are multiple calls to np.float128. This datatype is not available for windows machines. This results in continuous exceptions and constant rejection of models.
Suggest changing to "np.longdouble". This will default to float128 on linux and float64 on windows, which is the best you can do.
Changes should be made to "utilities" line 670 and "inference1D" line 382, though there may be other instances.
there appears to be a problem with the plotDatapoints.py script when one or more soundings have failed to converge (burn in). In this case, you cannot plot any of the points and the following error shows up:
[pbedrosian@yeti-login20 plotting] python plotDatapoints.py ../Data/ ../results/1650.0.h5 --points 5
Traceback (most recent call last):
File "plotDatapoints.py", line 51, in
fids = LR.iDs[args.points]
IndexError: too many indices for array
[pbedrosian@yeti-login20 plotting] python plotDatapoints.py ../Data/ ../results/1650.0.h5
Traceback (most recent call last):
File "plotDatapoints.py", line 64, in
for fid in fids:
TypeError: iteration over a 0-d array
From CultPenguin.
Setting user defined bounds on the magnitude of the recovered parameter is not possible.
Add an additional uniform prior to the parameters.
docs link receiving 404 error, https://usgs.github.io/geobipy/
Is there an option to save all (or every N'th realizations) from the Markov chain?
I do not see these realizations on the hdf5 output file?
With self.reciprocateParameters = True or False, the plot is always conductivity in a geobipySerial run of fdem data
I followed your instruction in the Documentation, the error messages are:
(base) C:\geobipy-master>setup.py install
Traceback (most recent call last):
File "C:\geobipy-master\setup.py", line 25, in
long_description=readme(),
File "C:\geobipy-master\setup.py", line 18, in readme
return f.read()
UnicodeDecodeError: 'gbk' codec can't decode byte 0x93 in position 870: illegal multibyte sequence
when using the plotDataPoints.py script, is there a way to select the sounding to be plotted by either its fiducial value or by the northing or easting value?
I need to update geobipy since Pandas removed their fast moving average window!
GeoBIPy works with 0.19.2
Hi
When I import geobipy (fter a fresh install) I get the following:
(geobipy) tmeha@pop-os:~/space/PROGRAMMING/geobipy/documentation_source/source/examples/Models$ python
Python 3.8.1 (default, Jan 8 2020, 22:29:32)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import geobipy
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/mnt/space/PROGRAMMING/geobipy/geobipy/__init__.py", line 21, in <module>
from .src.base import customFunctions
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/base/customFunctions.py", line 5, in <module>
from ..classes.core import StatArray
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/classes/core/StatArray.py", line 12, in <module>
from ...base.HDF import hdfRead
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/base/HDF/hdfRead.py", line 6, in <module>
from ...classes.data.datapoint.FdemDataPoint import FdemDataPoint
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/classes/data/datapoint/FdemDataPoint.py", line 7, in <module>
from .EmDataPoint import EmDataPoint
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/classes/data/datapoint/EmDataPoint.py", line 1, in <module>
from .DataPoint import DataPoint
File "/mnt/space/PROGRAMMING/geobipy/geobipy/src/classes/data/datapoint/DataPoint.py", line 1, in <module>
from cached_property import cached_property
ModuleNotFoundError: No module named 'cached_property'
Am I missing something ?
Hi
I run geobipySerial on a remote machine, and I do not use the X-server.
If I use
self.plot = False
self.savePNG = True
then geobipySerial stops with an error when the png file is being created:
"_tkinter.TclError: couldn't connect to display "localhost:12.0"
Is it possible to use another backend with matplotlib in this case (when self.plot=False) that does not need an Xserver?
The code runs fine if I use
self.plot = False
self.savePNG = False
Also, it would be nice if the progress of geobipySerial could be written to the console. Perhaps this is already an option?
Have a great day
I'm trying to plot some single-site results from geobipy. I can create the 'full' site plot using:
R = LR.getResults(LR.iDs[point])
R.plot(forcePlot=True)
Is there an easy way to plot just the component pieces of this figure (say just the histogram plot and the layer-depth historgram?
posterior plot for data elevation shows just a single value for all locations along a line.
the link https://usgs.github.io/geobipy/ takes a person to a 404 page.
How do you include the DOI as a transparency on top of a line plot? I tried using:
LR.plotDoi(linewidth=0.1, alpha=0.6, percent=40)
but got the following error-
File "plotLine.py", line 63, in
LR.plotDoi(linewidth=0.1, alpha=0.6, percent=40)
File "/cxfs/projects/software/geobipy/geobipy/src/inversion/LineResults.py", line 462, in plotDoi
self.getDOI(percent)
File "/cxfs/projects/software/geobipy/geobipy/src/inversion/LineResults.py", line 209, in getDOI
self.doi[i] = self.depthGrid.cellCentres[zSize - iC]
AttributeError: 'NoneType' object has no attribute 'cellCentres'
I want to use the 'Agg' backend to make plots (such that I can create and save figures remotely, without access to an X-server), as I use a remote server to run GeoBiPy.
But, if I do
import matplotlib as mpl
mpl.use('Agg')
import matplotlib.pyplot as plt
from geobipy import LineResults
I get an error impoerting LineResults from geobipy:
(geobipy) tmeha@Arizona:~/$ python pl_results.py
Traceback (most recent call last):
File "pl_results.py", line 15, in <module>
from geobipy import LineResults
File "/mnt/PROGRAMMING/geobipy/geobipy/__init__.py", line 20, in <module>
from .src.base import customPlots
File "/mnt/PROGRAMMING/geobipy/geobipy/src/base/customPlots.py", line 2, in <module>
mpl.use('TkAgg')
File "/home/tmeha/anaconda3/envs/geobipy/lib/python3.7/site-packages/matplotlib/__init__.py", line 1391, in use
switch_backend(name)
File "/home/tmeha/anaconda3/envs/geobipy/lib/python3.7/site-packages/matplotlib/pyplot.py", line 222, in switch_backend
newbackend, required_framework, current_framework))
ImportError: Cannot load backend 'TkAgg' which requires the 'tk' interactive framework, as 'headless' is currently running
A simple trick to allow a user to select the backend is to comment out line 2 in /geobipy/src/base/customPlots.py:
# mpl.use('TkAgg')
add to comments here what the units are for hard parameter limits (i.e. resistivity, conductivity, log/linear)
Hi there! I have installed the serial version of the code. When I followed the instruction to install the parallel version or TDEM modeller, it seems that the tutorial is not for Windows system.
I ran into issues when running pip install -e .
for installing the serial version. The error seems to be related to mtforward1d_fortranmodule.c
, which was not generated during build through f2py
. I am not sure how to fix this. Any help will be much appreciated!
I am using python 3.7.1, below is the error message. Python 3.5/3.6 also produced the same error.
$ pip install -e .
Obtaining file:///Users/.../code/geobipy
Complete output from command python setup.py egg_info:
running egg_info
running build_src
build_src
building extension "geobipy.src.classes.forwardmodelling.fdemforward1d_fortran" sources
f2py options: []
adding 'build/src.macosx-10.7-x86_64-3.7/build/src.macosx-10.7-x86_64-3.7/geobipy/src/classes/forwardmodelling/fortranobject.c' to sources.
adding 'build/src.macosx-10.7-x86_64-3.7/build/src.macosx-10.7-x86_64-3.7/geobipy/src/classes/forwardmodelling' to include_dirs.
adding 'build/src.macosx-10.7-x86_64-3.7/geobipy/src/classes/forwardmodelling/fdemforward1d_fortran-f2pywrappers2.f90' to sources.
building extension "geobipy.src.classes.forwardmodelling.mtforward1d_fortran" sources
f2py options: []
f2py:> build/src.macosx-10.7-x86_64-3.7/geobipy/src/classes/forwardmodelling/mtforward1d_fortranmodule.c
IOError: [Errno 2] No such file or directory: 'geobipy/src/classes/forwardmodelling/mtforward1D_fortran/m_mtforward1D.f90'. Skipping file "geobipy/src/classes/forwardmodelling/mtforward1D_fortran/m_mtforward1D.f90".
error: f2py target file 'build/src.macosx-10.7-x86_64-3.7/geobipy/src/classes/forwardmodelling/mtforward1d_fortranmodule.c' not generated
Reading fortran codes...
Post-processing...
Post-processing (stage 2)...
Building modules...
----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /Users/.../code/geobipy/
Thanks!
The first time a HDF5 file is used to calculate opacity, store it as a variable such that it can simply be read in if present.
document somewhere allowed column header names. There is a list printed in the .out file when you get one wrong, but can we include them someplace else?
add a parameter such as self.thin in the userParameters file that is used to thin the Markov Chain results so that only every nth model is included in output calculations & statistics. For example, with if the chain is run for 400,000 iterations with thin=4, the output statistics would be based on every 4th model (100,000 total). Accept/reject rules apply to all models in the chain as the algorithm runs.
This might help to deal with some 'noisy' outputs where the chain gets stuck occasionally, and will be a step towards future multi-chain runs
Hi there! In the docs, it says:
The code and its processes have been documented in multiple ways. First we have the publication associated with this software release, the citation is below, and presents the application of this package to frequency and time domain electro-magnetic inversion.
But there's no citation... would you all provide a link to the published work? Thanks!
Note: I found this abstract
Hi
Is there an example of howto run geobipy from within python, without the need to run it from the command line?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.