Problems with pymc3

takes a lot of time (although accuracy seems ok)
unclear how the distance is computed
unclear how to use summary statistics
unclear how to use custom distance
unclear how to bring computation time down (maybe I don't need that much accuracy)

pyABC?

Setup python: https://www.uppmax.uu.se/support/user-guides/python-modules-guide/
launch job: sbatch -p core -n 1 -t 72:00:00 -A snic2019-8-227 launch.sh 3.0 0.0

python3 collect_data.py data/diyabc.json data/diyabc/*

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
Figs		Figs
.gitattributes		.gitattributes
.gitignore		.gitignore
CBM.py		CBM.py
ModelComparison-bootstrap.ipynb		ModelComparison-bootstrap.ipynb
ModelComparison.ipynb		ModelComparison.ipynb
NormalizeSumStats.ipynb		NormalizeSumStats.ipynb
Plot_expvalue_map.ipynb		Plot_expvalue_map.ipynb
Plot_lklh_map.ipynb		Plot_lklh_map.ipynb
Plot_logexpvalue_map.ipynb		Plot_logexpvalue_map.ipynb
Plot_posterior.ipynb		Plot_posterior.ipynb
Plot_prob_map.ipynb		Plot_prob_map.ipynb
Plot_utility_map.ipynb		Plot_utility_map.ipynb
WMM.py		WMM.py
collect_data.py		collect_data.py
diy_abc.py		diy_abc.py
extract_data.sh		extract_data.sh
extract_model_expvalue.py		extract_model_expvalue.py
extract_model_likelihood.py		extract_model_likelihood.py
extract_model_logerror.py		extract_model_logerror.py
extract_model_logexpvalue.py		extract_model_logexpvalue.py
extract_model_prob.py		extract_model_prob.py
extract_model_utility.py		extract_model_utility.py
generate_data.ipynb		generate_data.ipynb
generate_inference_commands.py		generate_inference_commands.py
generate_run_commands.py		generate_run_commands.py
launch_inference.sh		launch_inference.sh
models.py		models.py
reactionrates.py		reactionrates.py
readme.md		readme.md
run_inference.py		run_inference.py
smoldyn.py		smoldyn.py
smoldyn_rackham.py		smoldyn_rackham.py
start_inf.sh		start_inf.sh