DICE

Huasheng Nie (Ph.D. student at UCLA) and I (Ivo Welch) are helping William Nordhaus adapt and distribute the DICE 2023 model in a python version. Python has the advantage of being open source and known to a lot more researchers.

We are trying to make the model a little easier to modify. For example, the parameters have been broken into a configuration file --- like `basescenario.conf`. Thus, it is easy to consider alternative scenarios without editing code. It requires only editing this configuration file and rerunning dice with the updated version.

We are distributing the model as a gzipped tar file, for installation into a fresh ubuntu linux host. If you do not know about ubuntu, you can install it via's iso images as a "virtual self-contained machine," such as the user-friendly UTM program on macos.

You may first want to look at the README.html. This README file is also in the distribution, dice-latest.tgz (itself a symbolic link to the most recent version [version 0.2 in Jan 2024]).

In your fresh running ubuntu virtual machine, start by

        $ wget https://www.ivo-welch.info/research/dice/dice-latest.tgz
        $ tar xvfz dice-latest.tgz
        $ cd dice-*/
	$ more _README_*
      

Follow the steps in the README file (mostly `sudo sh install.sh`, then `python3 dice.py nordhaus-base.conf`. The results will then be in `nordhaus-base.conf_fig.pdf` drawn from `nordhaus-base.conf-result.csv` (for convenience also downloadable from this page).