Organization

The main model is implemented in C++ as a static library (.a). This implementation is in libolfsysm/src/olfsysm.cpp, with API declarations in libolfsysm/api/olfsysm.hpp. The model is a port of Ann Kennedy's model, with a few modifications.

R bindings to the implementation library are in bindings/R/olfsysm/, which is structured as an R package. The R interface is implemented in bindings/R/olfsysm/R/olfsysm.R. The functions there closely mirror the API functions documented in libolfsysm/api/olfsysm.hpp.

Installation

R bindings

First install the R packages xptr and Rcpp.

Once these are installed, make from the base directory will build everything and install the R package. If not run with sudo, a local R install directory must already exist.

Compilation will fail with g++ version 5.5.0, but works with g++ 7.4.0. If using Ubuntu 16.04, you may need to follow these instructions to update your g++.

Python bindings

pip install pybind11
pip install .

To see compilation errors, which can be helpful when modify the library code / bindings:

pip install -v .

Most recently tested using:

• pip 22.3.1
• pybind11==2.10.4
• Python 3.8.12
• Ubuntu 20.04

Example Run

R

library(olfsysm)

mp <- mk_modelparams()
load_hc_data(mp, "hc_data.csv")
rv <- mk_runvars(mp)

run_ORN_LN_sims(mp, rv)
run_PN_sims(mp, rv)
run_KC_sims(mp, rv, T)

kc_resp <- get_rvar(rv, "kc.responses")

Python

import olfsysm as osm
mp = osm.ModelParams()
osm.load_hc_data(mp, "hc_data.csv")
rv = osm.RunVars(mp)

osm.run_ORN_LN_sims(mp, rv)
osm.run_PN_sims(mp, rv)
osm.run_KC_sims(mp, rv, True)

kc_resp = rv.kc.responses