Use Hashmaps, vector preallocation and avoid cloning [Rust] (#250)

* get hashmap without match statements

* preallocate vectors

* avoid cloning

Cargo.toml

@@ -6,7 +6,7 @@ edition = "2021"
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 [lib]
 name = "rshgf"
-crate-type = ["cdylib"]
+crate-type = ["cdylib", "rlib"]
 path = "src/lib.rs"    # The source file of the target.
 
 [dependencies]

examples/exponential.rs

@@ -0,0 +1,22 @@
+use rshgf::model::Network;
+
+fn main() {
+
+    // initialize network
+    let mut network = Network::new();
+
+    // create a network with two exponential family state nodes
+    network.add_nodes(
+        "exponential-state",
+        None,
+        None,
+        None,
+        None
+    );
+
+    // belief propagation
+    let input_data = vec![1.0, 1.3, 1.5, 1.7];
+    network.set_update_sequence();
+    network.input_data(input_data);
+
+}

src/model.rs

@@ -1,6 +1,7 @@
 use std::collections::HashMap;
-use crate::{updates::observations::observation_update, utils::function_pointer::FnType};
+use crate::utils::function_pointer::FnType;
 use crate::utils::set_sequence::set_update_sequence;
+use crate::utils::beliefs_propagation::belief_propagation;
 use crate::utils::function_pointer::get_func_map;
 use pyo3::types::PyTuple;
 use pyo3::{prelude::*, types::{PyList, PyDict}};
@@ -129,95 +130,68 @@ impl Network {
         self.update_sequence = set_update_sequence(self);
     }
 
-    /// Single time slice belief propagation.
-    /// 
-    /// # Arguments
-    /// * `observations` - A vector of values, each value is one new observation associated
-    /// with one node.
-    pub fn belief_propagation(&mut self, observations_set: Vec<f64>) {
-
-        let predictions = self.update_sequence.predictions.clone();
-        let updates = self.update_sequence.updates.clone();
-
-        // 1. prediction steps
-        for (idx, step) in predictions.iter() {
-            step(self, *idx);
-        }
-
-        // 2. observation steps
-        for (i, observations) in observations_set.iter().enumerate() {
-            let idx = self.inputs[i];
-            observation_update(self, idx, *observations);
-        } 
-
-        // 3. update steps
-        for (idx, step) in updates.iter() {
-            step(self, *idx);
-        }
-    }
-
     /// Add a sequence of observations.
     /// 
     /// # Arguments
     /// * `input_data` - A vector of vectors. Each vector is a time series of observations
     /// associated with one node.
     pub fn input_data(&mut self, input_data: Vec<f64>) {
 
+        let n_time = input_data.len();
+        let predictions = self.update_sequence.predictions.clone();
+        let updates = self.update_sequence.updates.clone();
+
         // initialize the belief trajectories result struture
         let mut node_trajectories = NodeTrajectories {floats: HashMap::new(), vectors: HashMap::new()};
 
-        // add empty vectors in the floats hashmap
+        // preallocate empty vectors in the floats hashmap
         for (node_idx, node) in &self.attributes.floats {
             let new_map: HashMap<String, Vec<f64>> = HashMap::new();
             node_trajectories.floats.insert(*node_idx, new_map);
-            if let Some(attr) = node_trajectories.floats.get_mut(node_idx) {
-                for key in node.keys() {
-                    attr.insert(key.clone(), Vec::new());
-                }
+            let attr = node_trajectories.floats.get_mut(node_idx).expect("New map not found.");
+            for key in node.keys() {
+                attr.insert(key.clone(), Vec::with_capacity(n_time));
             }
-        }
-        // add empty vectors in the vectors hashmap
+            }
+
+        // preallocate empty vectors in the vectors hashmap
         for (node_idx, node) in &self.attributes.vectors {
             let new_map: HashMap<String, Vec<Vec<f64>>> = HashMap::new();
             node_trajectories.vectors.insert(*node_idx, new_map);
-            if let Some(attr) = node_trajectories.vectors.get_mut(node_idx) {
-                for key in node.keys() {
-                    attr.insert(key.clone(), Vec::new());
-                }
+            let attr = node_trajectories.vectors.get_mut(node_idx).expect("New vector map not found.");
+            for key in node.keys() {
+                attr.insert(key.clone(), Vec::with_capacity(n_time));
             }
-        }
+            }
+
 
         // iterate over the observations
         for observation in input_data {
 
             // 1. belief propagation for one time slice
-            self.belief_propagation(vec![observation]);
+            belief_propagation(self, vec![observation], &predictions, &updates);
 
             // 2. append the new beliefs in the trajectories structure
             // iterate over the float hashmap
             for (new_node_idx, new_node) in &self.attributes.floats {
                 for (new_key, new_value) in new_node {
                     // If the key exists in map1, append the vector from map2
-                    if let Some(old_node) = node_trajectories.floats.get_mut(&new_node_idx) {
-                        if let Some(old_value) = old_node.get_mut(new_key) {
-                            old_value.push(*new_value);
-                        }
+                    let old_node = node_trajectories.floats.get_mut(&new_node_idx).expect("Old node not found.");
+                    let old_value = old_node.get_mut(new_key).expect("Old value not found");
+                    old_value.push(*new_value);
                     }
                 }
-            }
+
             // iterate over the vector hashmap
             for (new_node_idx, new_node) in &self.attributes.vectors {
                 for (new_key, new_value) in new_node {
                     // If the key exists in map1, append the vector from map2
-                    if let Some(old_node) = node_trajectories.vectors.get_mut(&new_node_idx) {
-                        if let Some(old_value) = old_node.get_mut(new_key) {
-                            old_value.push(new_value.clone());
-                        }
+                    let old_node = node_trajectories.vectors.get_mut(&new_node_idx).expect("Old vector node not found.");
+                    let old_value = old_node.get_mut(new_key).expect("Old vector value not found.");
+                    old_value.push(new_value.clone());
                     }
                 }
             }
-        }
-
         self.node_trajectories = node_trajectories;
     }
 

src/updates/prediction_error/exponential.rs

@@ -11,25 +11,14 @@ use crate::math::sufficient_statistics;
 /// * `network` - The network after message passing.
 pub fn prediction_error_exponential_state_node(network: &mut Network, node_idx: usize) {
 
-    if let Some(floats_attributes) = network.attributes.floats.get_mut(&node_idx) {
-        if let Some(vectors_attributes) = network.attributes.vectors.get_mut(&node_idx) {
-            let mean = floats_attributes.get("mean");
-            let nus = floats_attributes.get("nus");
-            let xis = vectors_attributes.get("xis");
-            let new_xis = match (mean, nus, xis) {
-                (Some(mean), Some(nus), Some(xis)) => {
-                    let suf_stats = sufficient_statistics(mean);
-                    let mut new_xis = xis.clone();
-                    for i in 0..suf_stats.len() {
-                        new_xis[i] = new_xis[i] + (1.0 / (1.0 + nus)) * (suf_stats[i] - xis[i]);
-                    }
-                    new_xis
-                    }
-                    _ => Vec::new(),
-                };
-                if let Some(xis) = vectors_attributes.get_mut("xis") {
-                    *xis = new_xis; // Modify the value directly
-                } 
-            }
-        }
-    }
+    let floats_attributes = network.attributes.floats.get_mut(&node_idx).expect("No floats attributes");
+    let vectors_attributes = network.attributes.vectors.get_mut(&node_idx).expect("No vector attributes");
+    let mean = floats_attributes.get("mean").expect("Mean not found");
+    let nus = floats_attributes.get("nus").expect("Nus not found");
+    let xis = vectors_attributes.get_mut("xis").expect("Xis not found");
+
+    let suf_stats = sufficient_statistics(mean);
+    for i in 0..suf_stats.len() {
+        xis[i] = xis[i] + (1.0 / (1.0 + nus)) * (suf_stats[i] - xis[i]);
+    }    
+}

src/utils/beliefs_propagation.rs

@@ -0,0 +1,25 @@
+use crate::{utils::function_pointer::FnType, model::Network, updates::observations::observation_update};
+
+/// Single time slice belief propagation.
+/// 
+/// # Arguments
+/// * `observations` - A vector of values, each value is one new observation associated
+/// with one node.
+pub fn belief_propagation(network: &mut Network, observations_set: Vec<f64>, predictions: & Vec<(usize, FnType)>, updates: & Vec<(usize, FnType)>) {
+
+    // 1. prediction steps
+    for (idx, step) in predictions.iter() {
+        step(network, *idx);
+    }
+
+    // 2. observation steps
+    for (i, observations) in observations_set.iter().enumerate() {
+        let idx = network.inputs[i];
+        observation_update(network, idx, *observations);
+    } 
+
+    // 3. update steps
+    for (idx, step) in updates.iter() {
+        step(network, *idx);
+    }
+}

src/utils/mod.rs

@@ -1,2 +1,3 @@
 pub mod set_sequence;
-pub mod function_pointer;
+pub mod function_pointer;
+pub mod beliefs_propagation;