Merge remote-tracking branch 'origin' into SAIL_316_snapshot_date

Author: AnneONS

helpers/regression_estimation.py

@@ -12,18 +12,18 @@
 
 sys.path.append("D:/programming_projects/research-and-development")
 #%%
-from src.utils.local_file_mods import read_local_csv as read_csv
-from src.utils.local_file_mods import write_local_csv as write_csv
+from src.utils.local_file_mods import rd_read_csv as read_csv
+from src.utils.local_file_mods import rd_write_csv as write_csv
 
 #%% Configuration settings
 
 # Input folder and file names
-in_fol = "D:/data/res_dev/regression\input/"
-in_file_old = "full_estimation_qa_2024-02-02_v106.csv"
-in_file_new = "full_estimation_qa_2024-02-02_v107_TEST.csv"
+in_fol = "R:/BERD Results System Development 2023/DAP_emulation/2023_surveys/BERD/08_estimation/estimation_qa/"
+in_file_old = "2023_full_estimation_qa_26-01-12_v1744.csv"
+in_file_new = "2023_full_estimation_qa_26-01-12_v1745.csv"
 
 # Output folder and file
-out_fol = "D:/data/res_dev/regression/output/"
+out_fol = "D:/data/res_dev/regression/"
 out_file = "merged.csv"
 
 # Columns to select

src/estimation/apply_weights.py

@@ -30,20 +30,17 @@ def apply_weights(
     employment_cols = get_all_wanted_columns(config, "employment")
     all_numeric_cols = get_all_wanted_columns(config, "estimation")
 
-    # if the dataframe is for QA output, create new columns with the weights applied.
     if for_qa:
         for col in all_numeric_cols:
-            df[f"{col}_estimated"] = round(df[col] * df["a_weight"], round_val)
+            df[f"{col}_estimated"] = df[col] * df["a_weight"]
         for col in employment_cols:
-            df[f"{col}_estimated"] = round(df[f"{col}_estimated"] * df["g_weight"], 0)
-
-    # otherwise, apply the weights directly to the existing columns
+            df[f"{col}_estimated"] = df[f"{col}_estimated"] * df["g_weight"]
     else:
         for col in all_numeric_cols:
-            df[col] = round(df[col] * df["a_weight"], round_val)
+            df[col] = df[col] * df["a_weight"]
         for col in employment_cols:
-            df[col] = round(df[col] * df["g_weight"], 0)
+            df[col] = df[col] * df["g_weight"]
 
-        df = calc_totals(df, config, "employment", 0)
+    df = calc_totals(df, config, "employment", 0)
 
     return df

src/utils/breakdown_validation.py

@@ -187,7 +187,11 @@ def calc_totals(
     for key, columns in equals_checks.items():
         if len(columns) == 1:
             continue
-        df[columns[-1]] = round(df[columns[:-1]].sum(axis=1), round_value)
+        summed = df[columns[:-1]].sum(axis=1)
+        if round_value > 0:
+            df[columns[-1]] = round(summed, round_value)
+        else:
+            df[columns[-1]] = summed
     return df