NCGR¶

NCGR is a Python package for calibrating probabilistic forecasts of ice-free dates and freeze-up dates using Non-homogenous Censored Gaussian Regression (NCGR) [1]. For plug-and-play execution of NCGR, please see:

Performing NCGR on a full forecast field

Installation¶

pip install NCGR

Simple Usage¶

from NCGR import ncgr

ncgr.ncgr_fullfield(fcst_netcdf, hc_netcdf, obs_netcdf, out_netcdf, event,
                    a, b, model_dict, obs_dict,
                    clim_netcdf=clim_netcdf)

Demos with Jupyter Notebook¶

Download notebooks and sample data

Documentation¶

NCGR
- ncgr
- dcnorm
- sitdates

Reference¶

[1] Dirkson, A., B. Denis., M.,Sigmond., & Merryfield, W.J. (2020). Development and Calibration of Seasonal Probabilistic Forecasts of Ice-free Dates and Freeze-up Dates. Weather and Forecasting. doi:10.1175/WAF-D-20-0066.1.

Indices and tables¶