L. Markowsky, Jessica Scheick
Chemical constituents measured in ice cores are often visualized in a standard format, with figures produced by proprietary or closed tools. Such tools present an extra step that requires a context switch in the process of analyzing and visualizing results. Python and its many machine learning, numerical computation, and visualization libraries together with Jupyter notebooks offer a unified alternative that permits researchers to analyze and visualize ice core datasets in a single, highly-integrated ecosystem. This Jupyter notebook uses exclusively open-source Python libraries (specifically NumPy, Matplotlib, and Pandas) to create a readily reproducible, publication quality figure. We demonstrate the utility of the notebook as a template for ice core researchers engaging in open science by recreating multiple figures published by Schupbach et al. in Nature Communications, 16 April 2018. This notebook provides a step-by-step guide to systematically recreate two figures using Schupbach's original data.
This notebook may be run on a local machine under Python (>=3.8) with the following minimal packages:
| Library | Min Version | Description |
|---|---|---|
| NumPy | 1.17.4 | Efficient, multi-dimensional array operations |
| Pandas | 0.25.3 | Data preparation and cleaning; SQL-like data manipulation |
| Matplotlib | 3.1.2 | Python plotting library with low-level control |
| Seaborn | 0.10.0 | Python visualization |
| OpenPyXL | 3.0.3 | Python library to read/write Excel files |