diff --git a/src/MSDpostprocess/models.py b/src/MSDpostprocess/models.py
index b4d78e8..1e553f4 100644
--- a/src/MSDpostprocess/models.py
+++ b/src/MSDpostprocess/models.py
@@ -8,14 +8,11 @@
 # import shelve
 import re
 from functools import cache
-import warnings
 
 import pickle
 import numpy as np
 import pandas as pd
-with warnings.catch_warnings():
-    warnings.filterwarnings("ignore", message="brainpy")
-    from brainpy import isotopic_variants
+from brainpy import isotopic_variants
 import statsmodels.api as sm
 from lineartree import LinearTreeClassifier
 from sklearn.preprocessing import StandardScaler
@@ -57,7 +54,6 @@ def fit(self, data):
         data = data.copy()
         data = data[np.isfinite(data['label'])]
         data = self.preprocess(data.copy())
-        # self.classifier = LogisticRegression(solver = 'liblinear')
         self.classifier = CalibratedClassifierCV(GBC(), ensemble=False)
         self.classifier.fit(data[self.features], data['label'])
         self.logs.info(f'Fit {self.model}')
@@ -172,7 +168,7 @@ def rt_error(self, subset):
                             frac = self.lowess_frac,
                             it = 3,
                             xvals = subset['Average Rt(min)'])
-        rt_error = subset['Reference RT'].to_numpy() - regression
+        subset['rt_error'] = subset['Reference RT'].to_numpy() - regression
         #record regression for QC purposes
         file = next(f for f in subset['file'])
         self.rt_observed[file] = subset['Average Rt(min)']
@@ -180,7 +176,7 @@ def rt_error(self, subset):
         self.rt_predictions[file] = regression
         self.rt_calls[file] = subset['call']
         self.logs.debug(f'RT regression fit for {file} used {np.sum(subset["call"])} observations')
-        return rt_error    
+        return subset    
 
     def correct_data(self, data):
         #identify high confidence subset for correction
@@ -188,9 +184,11 @@ def correct_data(self, data):
         data['call'] = self.predict(scores)
         
         #build lowess regressions
-        rt_error = data.groupby('file')[data.columns].apply(self.rt_error)
-        data['rt_error'] = [val for file in rt_error for val in file]
-        data = data.drop(columns = ['call'])
+        subsets = []
+        for file in set(data['file']):
+            subsets.append(self.rt_error(data[data['file'] == file]))
+        data = pd.concat(subsets)
+        data.drop(columns = ['call'])
         
         self.logs.info('RT correction has been applied.')
         return data