diff --git a/workflow/scripts/weighted_distances.py b/workflow/scripts/weighted_distances.py
index e7ea083..33538df 100644
--- a/workflow/scripts/weighted_distances.py
+++ b/workflow/scripts/weighted_distances.py
@@ -95,14 +95,19 @@ def calculate_distance(positions: List[int], sample1_name: str, sample2_name: st
                     for pos in positions])
 
 
-def buildm(df: pd.DataFrame, sample_names: list, reference: Seq) -> pd.DataFrame:
-    # TODO
-    # Pre-compute one-sample measurements
-    logging.debug(f"Caching computations")
-    cache = {}
+def build_cache(df: pd.DataFrame, reference: Seq):
+    cache = {"freq": {}, "hz": {}}
     for (sample, pos), site_df in df.groupby(["REGION", "POS"]):
+        if sample not in cache["freq"]:
+            cache["freq"][sample] = {}
+        if sample not in cache["hz"]:
+            cache["hz"][sample] = {}
         cache["freq"][sample][pos] = get_frequencies_in_position(site_df, pos, reference)
         cache["hz"][sample][pos] = heterozygosity(cache[sample][pos])
+    return cache
+
+
+def buildm(df: pd.DataFrame, sample_names: list, cache: dict) -> pd.DataFrame:
     # Compute matrix
     logging.debug(f"Filling distance matrix")
     nsamples = len(sample_names)
@@ -133,8 +138,12 @@ def main():
     variant_table = pd.concat([input_table, ancestor_table], ignore_index=True)
     sample_names = snakemake.params.samples + [reference.id]
     
+    # Pre-compute one-sample measurements
+    logging.info(f"Caching computations")
+    cache = build_cache(df, ancestor.seq)
+
     logging.info(f"Calculating matrix")
-    df = buildm(variant_table, sample_names, ancestor.seq)
+    df = buildm(variant_table, sample_names, cache)
     
     logging.info("Writing results")
     df.to_csv(snakemake.output.distances)