Fix/remove append_default_point to improve accuracy (#142)

open-spaced-repetition · Dec 21, 2023 · f45f46b · f45f46b
1 parent 8ff6ccf
commit f45f46b
Showing 1 changed file with 3 additions and 18 deletions.
diff --git a/src/pre_training.rs b/src/pre_training.rs
@@ -106,14 +106,6 @@ fn loss(
     logloss + l1
 }
 
-fn append_default_point(data: &mut Vec<AverageRecall>, default_s0: f32) {
-    data.push(AverageRecall {
-        delta_t: default_s0,
-        recall: 0.9,
-        count: 16.0,
-    });
-}
-
 const MIN_INIT_S0: f32 = 0.1;
 const MAX_INIT_S0: f32 = 100.0;
 
@@ -127,11 +119,7 @@ fn search_parameters(
     for (first_rating, data) in &mut pretrainset {
         let r_s0_default: HashMap<u32, f32> = R_S0_DEFAULT_ARRAY.iter().cloned().collect();
         let default_s0 = r_s0_default[first_rating];
-
-        append_default_point(data, default_s0);
-
         let delta_t = Array1::from_iter(data.iter().map(|d| d.delta_t));
-
         let recall = {
             // Laplace smoothing
             // (real_recall * n + average_recall * 1) / (n + 1)
@@ -140,7 +128,6 @@ fn search_parameters(
             let n = data.iter().map(|d| d.count).sum::<f32>();
             (real_recall * n + average_recall) / (n + 1.0)
         };
-
         let count = Array1::from_iter(data.iter().map(|d| d.count));
 
         let mut low = MIN_INIT_S0;
@@ -340,7 +327,7 @@ mod tests {
             ],
         )]);
         let actual = search_parameters(pretrainset, 0.9);
-        Data::from([*actual.get(&4).unwrap()]).assert_approx_eq(&Data::from([1.230_132_3]), 4);
+        Data::from([*actual.get(&4).unwrap()]).assert_approx_eq(&Data::from([0.944_284]), 4);
     }
 
     #[test]
@@ -349,10 +336,8 @@ mod tests {
         let items = anki21_sample_file_converted_to_fsrs();
         let average_recall = calculate_average_recall(&items);
         let pretrainset = split_data(items, 1).0;
-        Data::from(pretrain(pretrainset, average_recall).unwrap()).assert_approx_eq(
-            &Data::from([0.956_017_43, 1.694_406_5, 3.998_023_5, 8.268_223]),
-            4,
-        )
+        Data::from(pretrain(pretrainset, average_recall).unwrap())
+            .assert_approx_eq(&Data::from([1.001_276, 1.811_072, 4.405_640, 8.532_001]), 4)
     }
 
     #[test]