This summer internship position will be exploring new diagnostic tools for users to visualize time series information of 3D Spatial data using Python. Currently all of are scripts are written in Matlab, and we would are aiming to create a suite of diagnostic tools that are easy for users to integrate into their own work flow, and make plots that are intuitive and comprehensive.
The typical set-up is a "twin experiment", where you
- specify a
- dynamic model*
- observational model*
- use these to generate a synthetic
- "truth"
- and observations thereof*
- assess how different DA methods perform in estimating the truth, given the above starred (*) items.
Prerequisite: python3.5+ with
scipy, matplotlib, pandas.
This is all comes with anaconda
by default.
For the tutorials, you will also need
jupyter and the markdown package.
It is also recommended to install tqdm (e.g. pip install tqdm).
| Model | Linear? | Phys.dim. | State len | # Lyap≥0 | Thanks to |
|---|
Many
- Visualizations
- Diagnostics
- Tools to manage and display experimental settings and stats
- Reorganize file structure
- Turn into package?
- Simplify time management?
- Use pandas for stats time series?
- Complete QG