search

nansencenter / DAPPER

Data Assimilation with Python: a Package for Experimental Research
https://nansencenter.github.io/DAPPER
MIT License
341 stars 119 forks source link

Finding parameters of Lorenz63 #98

Closed sgrzegorz closed 1 year ago

sgrzegorz commented 2 years ago

How can I use Dapper to find parameters of Lorenz63? 1) Lets say the parameters are at the begining:

sig = 10.0
rho = 28.0
beta = 8.0 / 3

2) Then I use xx = simulator(x0, k=5*10**3, t0=0, dt=0.01) (like it is done in examples) and I have artificial ground truth data. It is easy for me to this moment. 3) Now I would like to forget the values of sig, rho, beta and estimate them again eg using 3DVar (or whatever else) 4) Could you please provide me a basic working script? Something similar is done in example3a.py but I am a beginner and it is too complicated for me. I think that I should provide parameter ranges there isn't it? eg.: sig should be between [4 and 12] (because 4<10.0<12, rho should be eg. between [10 and 30] and beta eg. between [1 and 10]

sgrzegorz commented 2 years ago

Help :))))

patnr commented 2 years ago

You could try to adapt examples/param_estim.py to your purpose.

Make sure to read examples/basic_1.py and basic_2.py first. And peruse the documentation. I'm afraid we don't have the capacity to furnish you with a ready made script.