|
|
|
@ -1,6 +1,11 @@ |
|
|
|
|
#![allow(non_snake_case)] |
|
|
|
|
|
|
|
|
|
use csv; |
|
|
|
|
use duration_string::DurationString; |
|
|
|
|
use json; |
|
|
|
|
use std::collections::BTreeSet; |
|
|
|
|
use std::fs; |
|
|
|
|
use std::time::Instant; |
|
|
|
|
|
|
|
|
|
use reCTBN::parameter_learning::MLE; |
|
|
|
|
use reCTBN::params::DiscreteStatesContinousTimeParams; |
|
|
|
@ -17,10 +22,19 @@ use reCTBN::tools::RandomParametersGenerator; |
|
|
|
|
use reCTBN::tools::UniformGraphGenerator; |
|
|
|
|
use reCTBN::tools::UniformParametersGenerator; |
|
|
|
|
|
|
|
|
|
fn uniform_parameters_generator_right_densities_ctmp() -> (CtbnNetwork, Dataset) { |
|
|
|
|
fn uniform_parameters_generator_right_densities_ctmp( |
|
|
|
|
nodes_cardinality: usize, |
|
|
|
|
domain_cardinality: usize, |
|
|
|
|
density: f64, |
|
|
|
|
interval_s: f64, |
|
|
|
|
interval_f: f64, |
|
|
|
|
n_trajectories: u64, |
|
|
|
|
t_end: f64, |
|
|
|
|
sg_seed: Option<u64>, |
|
|
|
|
cg_seed: Option<u64>, |
|
|
|
|
tg_seed: Option<u64>, |
|
|
|
|
) -> (CtbnNetwork, Dataset) { |
|
|
|
|
let mut net = CtbnNetwork::new(); |
|
|
|
|
let nodes_cardinality = 20; |
|
|
|
|
let domain_cardinality = 3; |
|
|
|
|
for node in 0..nodes_cardinality { |
|
|
|
|
// Create the domain for a discrete node
|
|
|
|
|
let mut domain = BTreeSet::new(); |
|
|
|
@ -37,22 +51,41 @@ fn uniform_parameters_generator_right_densities_ctmp() -> (CtbnNetwork, Dataset) |
|
|
|
|
|
|
|
|
|
// Initialize the Graph Generator using the one with an
|
|
|
|
|
// uniform distribution
|
|
|
|
|
let mut structure_generator = UniformGraphGenerator::new(1.0 / 3.0, Some(7641630759785120)); |
|
|
|
|
let mut structure_generator = UniformGraphGenerator::new(density, sg_seed); |
|
|
|
|
|
|
|
|
|
// Generate the graph directly on the network
|
|
|
|
|
structure_generator.generate_graph(&mut net); |
|
|
|
|
|
|
|
|
|
// Initialize the parameters generator with uniform distributin
|
|
|
|
|
let mut cim_generator = UniformParametersGenerator::new(3.0..7.0, Some(7641630759785120)); |
|
|
|
|
let mut cim_generator = UniformParametersGenerator::new(interval_s..interval_f, cg_seed); |
|
|
|
|
|
|
|
|
|
// Generate CIMs with uniformly distributed parameters.
|
|
|
|
|
cim_generator.generate_parameters(&mut net); |
|
|
|
|
|
|
|
|
|
let dataset = trajectory_generator(&net, 300, 200.0, Some(30230423)); |
|
|
|
|
let dataset = trajectory_generator(&net, n_trajectories, t_end, tg_seed); |
|
|
|
|
|
|
|
|
|
return (net, dataset); |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
fn write_csv_record<I, T>(file_name: &str, record: I) -> std::io::Result<()> |
|
|
|
|
where |
|
|
|
|
I: IntoIterator<Item = T>, |
|
|
|
|
T: AsRef<[u8]>, |
|
|
|
|
{ |
|
|
|
|
let file = fs::OpenOptions::new() |
|
|
|
|
.write(true) |
|
|
|
|
.create(true) |
|
|
|
|
.append(true) |
|
|
|
|
.open(file_name) |
|
|
|
|
.unwrap(); |
|
|
|
|
let mut wtr = csv::Writer::from_writer(file); |
|
|
|
|
|
|
|
|
|
//let mut wtr = csv::Writer::from_path(file_name)?;
|
|
|
|
|
wtr.write_record(record)?; |
|
|
|
|
wtr.flush()?; |
|
|
|
|
Ok(()) |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
fn structure_learning_CTPC(net: CtbnNetwork, dataset: &Dataset) { |
|
|
|
|
// Initialize the hypothesis tests to pass to the CTPC with their
|
|
|
|
|
// respective significance level `alpha`
|
|
|
|
@ -67,6 +100,72 @@ fn structure_learning_CTPC(net: CtbnNetwork, dataset: &Dataset) { |
|
|
|
|
} |
|
|
|
|
|
|
|
|
|
fn main() { |
|
|
|
|
let (net, dataset) = uniform_parameters_generator_right_densities_ctmp(); |
|
|
|
|
let file_path = "./networks-settings.json"; |
|
|
|
|
let csv_path = "./results.csv"; |
|
|
|
|
eprintln!("Opening file {}...", file_path); |
|
|
|
|
let file_content = |
|
|
|
|
fs::read_to_string(file_path).expect("File not found! Check the README for instructions!"); |
|
|
|
|
let parsed_json = json::parse(&file_content).unwrap(); |
|
|
|
|
|
|
|
|
|
write_csv_record( |
|
|
|
|
csv_path, |
|
|
|
|
&[ |
|
|
|
|
"nodes_cardinality", |
|
|
|
|
"domain_cardinality", |
|
|
|
|
"density", |
|
|
|
|
"duration", |
|
|
|
|
], |
|
|
|
|
) |
|
|
|
|
.unwrap(); |
|
|
|
|
|
|
|
|
|
let benchmarks_cardinality = parsed_json.len(); |
|
|
|
|
for idx in 0..benchmarks_cardinality { |
|
|
|
|
eprintln!( |
|
|
|
|
"[{}/{}] Generating network...", |
|
|
|
|
idx + 1, |
|
|
|
|
benchmarks_cardinality |
|
|
|
|
); |
|
|
|
|
let (net, dataset) = uniform_parameters_generator_right_densities_ctmp( |
|
|
|
|
parsed_json[idx]["nodes_cardinality"].as_usize().unwrap(), |
|
|
|
|
parsed_json[idx]["domain_cardinality"].as_usize().unwrap(), |
|
|
|
|
parsed_json[idx]["density"].as_f64().unwrap(), |
|
|
|
|
parsed_json[idx]["interval_s"].as_f64().unwrap(), |
|
|
|
|
parsed_json[idx]["interval_f"].as_f64().unwrap(), |
|
|
|
|
parsed_json[idx]["n_trajectories"].as_u64().unwrap(), |
|
|
|
|
parsed_json[idx]["t_end"].as_f64().unwrap(), |
|
|
|
|
parsed_json[idx]["sg_seed"].as_u64(), |
|
|
|
|
parsed_json[idx]["cg_seed"].as_u64(), |
|
|
|
|
parsed_json[idx]["tg_seed"].as_u64(), |
|
|
|
|
); |
|
|
|
|
eprintln!( |
|
|
|
|
"[{}/{}] Structure learning CTPC...", |
|
|
|
|
idx + 1, |
|
|
|
|
benchmarks_cardinality |
|
|
|
|
); |
|
|
|
|
let start = Instant::now(); |
|
|
|
|
structure_learning_CTPC(net, &dataset); |
|
|
|
|
let duration = start.elapsed(); |
|
|
|
|
eprintln!( |
|
|
|
|
"[{}/{}] Strucure learned with CTPC in {:?}.", |
|
|
|
|
idx + 1, |
|
|
|
|
benchmarks_cardinality, |
|
|
|
|
duration |
|
|
|
|
); |
|
|
|
|
write_csv_record( |
|
|
|
|
csv_path, |
|
|
|
|
&[ |
|
|
|
|
parsed_json[idx]["nodes_cardinality"] |
|
|
|
|
.as_usize() |
|
|
|
|
.unwrap() |
|
|
|
|
.to_string(), |
|
|
|
|
parsed_json[idx]["domain_cardinality"] |
|
|
|
|
.as_usize() |
|
|
|
|
.unwrap() |
|
|
|
|
.to_string(), |
|
|
|
|
parsed_json[idx]["density"].as_f64().unwrap().to_string(), |
|
|
|
|
DurationString::from(duration).to_string(), |
|
|
|
|
], |
|
|
|
|
) |
|
|
|
|
.unwrap(); |
|
|
|
|
} |
|
|
|
|
} |
|
|
|
|