preliminary tests show that the refactorisation of a-LD-kNNi has impr…

…oved computational efficiency

src/imputation/adaptive_ld_knn_imputation.rs

@@ -3,8 +3,8 @@ use crate::gwas::pearsons_correlation;
 use ndarray::{prelude::*, Zip};
 use std::io;
 
-// For benchmarking
-use std::time::Instant;
+// // For benchmarking
+// use std::time::Instant;
 
 fn calculate_euclidean_distances(
     window_freqs: ArrayView2<f64>,
@@ -319,7 +319,7 @@ impl GenotypesAndPhenotypes {
                             } // Impute across pools with missing data
                         } // Impute if we have missing data
                     } // Iterate across alleles across loci within the window
-                    println!("window_freqs:\n{:?}", &window_freqs);
+                    // println!("window_freqs:\n{:?}", &window_freqs);
                 }, // Parallel processing across windows
             );
         // }

src/main.rs

@@ -121,7 +121,7 @@ struct Args {
     /// Imputation parameter, i.e. maximum number of top correlated loci within the window which are considered in linkage disequillibrium (LD) with the locus requiring imputation. The resulting loci will be used to calculate pairwise distances (adaptive if we have too much missing data in the window at which point we use all the loci within the window).
     #[clap(long, default_value_t = 10)]
     n_loci_to_estimate_distance: u64,
-    /// Imputation parameter, i.e. number of nearest neighbours from which the imputed weighted (weights based on distance from the pool requiring imputation) mean allele frequencies will be calculated from (adaptive if all k neighbours are also requiring imputation at the locus, at which point we increase k until at least one pool in non-missing at the locus).
+    /// Imputation parameter, i.e. number of nearest neighbours from which the imputed weighted (weights based on distance from the pool requiring imputation) mean allele frequencies will be calculated from.
     #[clap(long, default_value_t = 5)]
     k_neighbours: u64,
 }