Add in compensation for whether numerical grading scales include zero or not

multiscale_rt_critics.py

@@ -34,6 +34,7 @@ import os
 import sys
 import pandas
 import numpy
+import math
 from os.path import join as pjoin
 from datasets import Dataset
 from sklearn.model_selection import train_test_split
@@ -143,17 +144,38 @@ def np_round(arr):
     return (arr + 0.5).astype(numpy.int32)
 
 
-def common_denom_grades(group_df):
+def process_letter_grade_group(group_df):
+    group_df["includes_zero"] = False
+    group_df["multiplier"] = 1
+    group_df["non_neg_error"] = False
+    if group_df.iloc[0]["letter_implies_short"]:
+        group_df["label"] = SHORT_LETTER_SCALE.index(group_df.iloc[0]["review_score"])
+        group_df["scale_points"] = len(SHORT_LETTER_SCALE)
+    else:
+        group_df["label"] = LONG_LETTER_SCALE.index(group_df.iloc[0]["review_score"])
+        group_df["scale_points"] = len(LONG_LETTER_SCALE)
+    return group_df
+
+
+def process_includes_zero(group_df):
+    multiplier = group_df.iloc[0]["multiplier"]
+    includes_zero = any((label < multiplier for label in group_df["label"]))
+    group_df["includes_zero"] = includes_zero
+    if not includes_zero:
+        group_df["label"] -= multiplier
+        group_df["scale_points"] -= multiplier
+    return group_df
+
+
+def find_effective_nom_denom(group_df):
     if group_df.iloc[0]["is_any_letter"]:
-        group_df["multiplier"] = 1
-        group_df["non_neg_error"] = False
-        if group_df.iloc[0]["letter_implies_short"]:
-            group_df["num"] = SHORT_LETTER_SCALE.index(group_df.iloc[0]["review_score"]) + 1
-            group_df["denom"] = len(SHORT_LETTER_SCALE)
-        else:
-            group_df["num"] = LONG_LETTER_SCALE.index(group_df.iloc[0]["review_score"]) + 1
-            group_df["denom"] = len(LONG_LETTER_SCALE)
-        return group_df
+        return process_letter_grade_group(group_df)
+    else:
+        group_df = common_denom_grades(group_df)
+        return process_includes_zero(group_df)
+
+
+def common_denom_grades(group_df):
     denoms = numpy.empty(len(group_df), dtype=numpy.int32)
     for idx, num in enumerate(group_df["orig_num"]):
         frac = Fraction.from_float(num)
@@ -162,13 +184,14 @@ def common_denom_grades(group_df):
     group_df["multiplier"] = common_denom 
     num = common_denom * group_df["orig_num"].to_numpy()
     denom = common_denom * group_df["orig_denom"].to_numpy()
-    group_df["num"] = np_round(num)
-    group_df["denom"] = np_round(denom)
-    group_df["non_neg_error"] = (abs(group_df["num"] - num) >= 0.05) | (abs(group_df["denom"] - denom) >= 0.05)
+    group_df["label"] = np_round(num)
+    group_df["scale_points"] = np_round(denom)
+    group_df["non_neg_error"] = (abs(group_df["label"] - num) >= 0.05) | (abs(group_df["scale_points"] - denom) >= 0.05)
     return group_df
 
 
 def normalize_reviews(review_df):
+    print()
     # Drop unrated
     review_df = drop_unrated(review_df)
 
@@ -224,7 +247,7 @@ def normalize_reviews(review_df):
     print("unique grade/publisher combinations", working_review_df.groupby(["grade_type", "publisher_name"]).ngroups)
 
     # Now we can find common denominators on a (publisher, grade type) combination basis
-    working_review_df = working_review_df.groupby(["publisher_name", "grade_type"], group_keys=False).apply(common_denom_grades)
+    working_review_df = working_review_df.groupby(["publisher_name", "grade_type"], group_keys=False).apply(find_effective_nom_denom)
     working_review_df = drop_because(working_review_df, working_review_df["multiplier"] > 500, "multiplier > 500")
     assert working_review_df["non_neg_error"].sum() == 0
     
@@ -232,6 +255,10 @@ def normalize_reviews(review_df):
     print("non-neg error count", working_review_df["non_neg_error"].sum())
     print("multipliers")
     print(working_review_df["multiplier"].value_counts())
+    print("includes_zero")
+    print(working_review_df["includes_zero"].value_counts())
+    print("grade breakdown")
+    print(working_review_df.value_counts(["grade_type", "multiplier", "includes_zero", "scale_points"]))
 
     # TODO: Add back in rare review_scores dropped at the beginning when they
     # are compatible with some common denominator + grade type from the same
@@ -274,7 +301,7 @@ def split_dfs(df):
         group_cols["publisher_name"].append(publisher_name)
         group_cols["grade_type"].append(grade_type)
         group_cols["group_id"].append(group_id)
-        group_cols["scale_points"].append(group_df.iloc[0]["denom"])
+        group_cols["scale_points"].append(group_df.iloc[0]["scale_points"])
         group_id += 1
 
     for publisher_name, grade_type, group_df in split_groups:
@@ -331,8 +358,8 @@ NORMAL_FEATURES = datasets.Features({
     "grade_type": datasets.Value("string"),
     "orig_num": datasets.Value("float"),
     "orig_denom": datasets.Value("float"),
-    "num": datasets.Value("uint8"),
-    "denom": datasets.Value("uint8"),
+    "label": datasets.Value("uint8"),
+    "scale_points": datasets.Value("uint8"),
     "multiplier": datasets.Value("uint8"),
     "group_id": datasets.Value("uint32"),
 })