Add (json?) file specifying inference params etc.

[hyanwong]

We currently say

The version numbers of inference software, the specific inference parameters needed to reproduce the ARGs, and (where necessary) functions to convert outputs to tskit format, are available from our GitHub repository

I think it would help if we created a single (json?) file with the version numbers and parameters used - we could then simply read this file in the makefile for the tools and when running them in make_example_outputs.py

[hyanwong]

Actually, something like this (rather than JSON) seems to work fine, and can be parsed easily using the built-in python config parser, as well as simply in a Makefile

# Params for Drosophila inference
NE = 1e6
# Values below from stdpopsim
MU=5.49e-09
RHO = 8.4e-09

KWARG_GITREPO = https://github.com/a-ignatieva/kwarg
KWARG_VERSION = v1.0
# don't allow sequencing errors or recurrent mutations; set seed to 9999
KWARG_PARAMS = -S-1 -M-1 -Z9999

ARGWEAVER_GITREPO = https://github.com/CshlSiepelLab/argweaver
# ARGWEAVER doesn't have a tagged release, so use the latest as of Mar 2022
ARGWEAVER_VERSION = 7191343c1d43c02de04e4e0a150d77b0612728c6
ARGWEAVER_PARAMS = --smc-prime --popsize %(NE)s --mutrate %(MU)s --recombrate %(RHO)s --randseed 111 --iters 3 --sample-step 10000 --no-compress-output

# ...