Python Package for the Analysis of Paleoclimate Data
Table of contents
- What is it?
- Version Information
- Quickstart Guide
- Known Issues
- Further information
Pyleoclim is a Python package primarily geared towards the analysis and visualization of paleoclimate data. Such data often come in the form of timeseries with missing values and age uncertainties, so the package includes several low-level methods to deal with these issues, as well as high-level methods that re-use those within scientific workflows.
High-level modules assume that data are stored in the Linked Paleo Data (LiPD) format and makes extensive use of the LiPD utilities. Low-level modules are primarily based on NumPy arrays or Pandas dataframes, so Pyleoclim contains a lot of timeseries analysis code (e.g. spectral analysis, singular spectrum analysis, wavelet analysis, correlation analysis) that can apply to these more common types as well. See the example folder for details.
The package is aware of age ensembles stored via LiPD and uses them for time-uncertain analyses very much like GeoChronR.
- plotting maps, timeseries, and basic age model information
- paleo-aware correlation analysis (isopersistent, isospectral and classical t-test)
- weighted wavelet Z transform (WWZ)
- age modeling through Bchron
- paleo-aware singular spectrum analysis (AR(1) null eigenvalue identification, missing data)
- spectral analysis (Multi-Taper Method, Lomb-Scargle)
- cross-wavelet analysis
- index reconstruction
- climate reconstruction
- ensemble methods for most of the above
If you have specific requests, please contact firstname.lastname@example.org
Python v3.6+ is required.
To install Pyleoclim, first install numpy and Cartopy through Anaconda (
conda install numpy
conda install -c conda-forge cartopy
Then install pyleoclim via
pip install pyleoclim
Note that the
pip command line above will trigger the installation of (most of) the dependencies,
as well as the local compilation of the Fortran code for WWZ with the GNU Fortran compiler
If you have the Intel's Fortran compiler
ifort installed, then further accerlation for WWZ could be
achieved by compiling the Fortran code with
ifort, and below are the steps:
- download the source code, either via
git cloneor just download the .zip file
setup.pyby commenting out the line of
gfortran, and use the line below for
python setup.py build_ext --fcompiler=intelem && python setup.py install
Some functionalities require R.
0.4.10: Support local compilation of the Fortran code for WWZ; precompiled .so files have been removed.
0.4.9: Major bug fixes; mapping module based on cartopy; compatibility with latest numpy package
0.4.8: Add support of f2py WWZ for Linux
0.4.7: Update to coherence function
0.4.6: Fix an issue when copying the .so files
0.4.5: Update to setup.py to include proper .so file according to version
0.4.4: New fix for .so issue
0.4.3: New fix for .so issue
0.4.2: Fix issue concerning download of .so files
0.4.1: Fix issues with tarball
0.4.0: New functionalities: map nearest records by archive type, plot ensemble time series, age modelling through Bchron
0.3.1: New functionalities: segment a timeseries using a gap detection criteria, update to summary plot to perform spectral analysis
0.3.0: Compatibility with LiPD 1.3 and Spectral module added 0.2.5: Fix error on loading (Looking for Spectral Module)
0.2.4: Fix load error from init
0.2.3: Freeze LiPD version to 1.2 to avoid conflicts with 1.3
0.2.2: Change progressbar to tqdm and add standardization function
0.2.1: Update package requirements
0.2.0: Restructure the package so that the main functions can be called without the use of a LiPD files and associated timeseries objects.
0.1.4: Rename function using camel case and consistency with LiPD utilities version 0.1.8.5
0.1.3: Compatible with LiPD utilities version 0.1.8.5. Function openLiPD() renamed openLiPDs()
0.1.2: Compatible with LiPD utilities version 0.1.8.3. Uses basemap instead of cartopy
0.1.1: Freezes the package prior to version 0.1.8.2 of LiPD utilities
Wait for installation to complete, then:
3a. Import the package into your favorite Python environment (we recommend the use of Spyder, which comes standard with the Anaconda package)
3b. Use Jupyter Notebook to go through the tutorial contained in the
PyleoclimQuickstart.ipynbNotebook, which can be downloaded here. The folder also contains a collection of LiPD files. More LiPD files available here.
Help with functionalities can be found in the Documentation.
- LiPD 0.2.7
- pandas v0.25.0
- numpy v1.16.4
- matplotlib v3.1.0
- Cartopy v1.17.0
- scipy v1.3.1
- statsmodel v0.8.0
- seaborn 0.9.0
- scikit-learn 0.21.3
- tqdm 4.33.0
- pathos 0.2.4
- rpy2 3.0.5
The installer will automatically check for the needed updates.
- Some of the packages supporting Pyleoclim do not have a build for Windows
- Known issues with proj4 v5.0-5.1, make sure your environment is set up with v5.2
Please report issues to email@example.com
The project is licensed under the GNU Public License. Please refer to the file call license. If you use the code in publications, please credit the work using this citation.
This material is based upon work supported by the National Science Foundation under Grant Number ICER-1541029. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the National Science Foundation.
This research is funded in part by JP Morgan Chase & Co. Any views or opinions expressed herein are solely those of the authors listed, and may differ from the views and opinions expressed by JP Morgan Chase & Co. or its affilitates. This material is not a product of the Research Department of J.P. Morgan Securities LLC. This material should not be construed as an individual recommendation of for any particular client and is not intended as a recommendation of particular securities, financial instruments or strategies for a particular client. This material does not constitute a solicitation or offer in any jurisdiction.