Merge pull request #15 from bzhanglab/dev

use par_iter for sums, improve speed by 15%

Cargo.toml

@@ -11,9 +11,9 @@ edition = "2021"
 [dependencies]
 bincode = "1.3.3"
 clap = { version = "4.4.6", features = ["derive"] }
-owo-colors = {version = "3.5.0", features = ["supports-colors"]}
-serde = {version="1.0.188", features=["std", "derive"]}
-webgestalt_lib = {path = "webgestalt_lib"}
+owo-colors = { version = "3.5.0", features = ["supports-colors"] }
+serde = { version = "1.0.188", features = ["std", "derive"] }
+webgestalt_lib = { path = "webgestalt_lib" }
 
 [profile.release]
 opt-level = 3

webgestalt_lib/src/methods/gsea.rs

@@ -4,7 +4,6 @@ use ahash::AHashSet;
 use rand::prelude::SliceRandom;
 use rand::SeedableRng;
 use rayon::prelude::*;
-use std::sync::{Arc, Mutex};
 
 /// Parameters for GSEA
 #[derive(Clone)]
@@ -202,8 +201,8 @@ fn analyte_set_p(
             .collect();
         let up_len = up.len();
         let down_len = down.len();
-        let up_avg: f64 = up.iter().sum::<f64>() / (up_len as f64 + 0.000001) + 0.000001; // up average
-        let down_avg: f64 = down.iter().sum::<f64>() / (down_len as f64 + 0.000001) - 0.000001; // down average
+        let up_avg: f64 = up.par_iter().sum::<f64>() / (up_len as f64 + 0.000001) + 0.000001; // up average
+        let down_avg: f64 = down.par_iter().sum::<f64>() / (down_len as f64 + 0.000001) - 0.000001; // down average
         let mut nes_es: Vec<f64> = up.par_iter().map(|x| x / up_avg).collect(); // get all normalized scores for up
         nes_es.extend(down.par_iter().map(|x| -x / down_avg).collect::<Vec<f64>>()); // extend with down scores
         let norm_es: f64 = if real_es >= 0_f64 {

webgestalt_lib/src/methods/multiomics.rs

@@ -22,7 +22,7 @@ pub enum MetaAnalysisMethod {
 }
 
 pub enum AnalysisType {
-    /// Gene Set Enrichment Analysips
+    /// Gene Set Enrichment Analysis
     GSEA,
     /// Over-representation Analysis
     ORA,
@@ -49,12 +49,12 @@ pub enum NormalizationMethod {
     None,
 }
 
-/// Run a multiomics analysis, using iehter the max/mean median ratio or a typical meta analysis
+/// Run a multiomics analysis, using either the max/mean median ratio or a typical meta analysis
 /// method
 ///
 /// # Parameters
 ///
-/// - `jobs` - A [`Vec<GSEAJob>`] containing all of the seperates 'jobs' or analysis to combine
+/// - `jobs` - A [`Vec<GSEAJob>`] containing all of the separates 'jobs' or analysis to combine
 /// - `method` - A [`MultiOmicsMethod`] enum detailing the analysis method to combine the runs together (meta-analysis, mean median ration, or max median ratio).
 ///
 /// # Returns

webgestalt_lib/src/methods/ora.rs

@@ -2,7 +2,6 @@ use crate::{readers::utils::Item, stat};
 use ahash::AHashSet;
 use rayon::prelude::*;
 use statrs::distribution::{DiscreteCDF, Hypergeometric};
-use std::sync::{Arc, Mutex};
 
 #[derive(Clone)]
 pub struct ORAConfig {

webgestalt_lib/tests/gsea_tests.rs

@@ -1,4 +1,4 @@
-use pretty_assertions::{assert_eq, assert_ne};
+use pretty_assertions::assert_eq;
 use statrs::assert_almost_eq;
 use webgestalt_lib;
 

webgestalt_lib/tests/ora_tests.rs

@@ -9,7 +9,7 @@ fn ora() {
         "data/genelist.txt".to_owned(),
         "data/reference.txt".to_owned(),
     );
-    let gmtcount = gmt.len();
+    let gmtcount: usize = gmt.len();
     let x: Vec<webgestalt_lib::methods::ora::ORAResult> =
         webgestalt_lib::methods::ora::get_ora(&gene_list, &reference, gmt, ORAConfig::default());
     let res = x.iter().find(|x| x.set == "GO:2000147").unwrap();