diff --git a/ptype/callbacks.py b/ptype/callbacks.py
index f19be40..cefd047 100644
--- a/ptype/callbacks.py
+++ b/ptype/callbacks.py
@@ -5,7 +5,7 @@
     CSVLogger,
     EarlyStopping,
 )
-from evml.keras.models import calc_prob_uncertainty
+from mlguess.keras.models import calc_prob_uncertainty
 from tensorflow.python.keras.callbacks import ReduceLROnPlateau
 from sklearn.metrics import precision_recall_fscore_support, roc_auc_score
 from hagelslag.evaluation.ProbabilityMetrics import DistributedROC
diff --git a/ptype/models.py b/ptype/models.py
index ddd5f13..e1efdaf 100644
--- a/ptype/models.py
+++ b/ptype/models.py
@@ -13,8 +13,8 @@
 
 from imblearn.under_sampling import RandomUnderSampler
 from imblearn.tensorflow import balanced_batch_generator
-from evml.keras.losses import DirichletEvidentialLoss
-from evml.keras.callbacks import ReportEpoch
+from mlguess.keras.losses import DirichletEvidentialLoss
+from mlguess.keras.callbacks import ReportEpoch
 
 
 logger = logging.getLogger(__name__)
diff --git a/ptype/qc.py b/ptype/qc.py
index d8b88ad..6d9993d 100644
--- a/ptype/qc.py
+++ b/ptype/qc.py
@@ -8,7 +8,6 @@
 import xarray as xr
 import metpy.calc
 from metpy.units import units
-import pandas as pd
 import geopandas as gpd
 from shapely.geometry import Point
 import cartopy.io.shapereader as shpreader
diff --git a/ptype/reliability.py b/ptype/reliability.py
index 657adac..080de89 100644
--- a/ptype/reliability.py
+++ b/ptype/reliability.py
@@ -1,4 +1,3 @@
-import os
 import numpy as np
 import matplotlib.pyplot as plt
 
diff --git a/ptype/seed.py b/ptype/seed.py
index e53901a..6323c5d 100644
--- a/ptype/seed.py
+++ b/ptype/seed.py
@@ -4,14 +4,15 @@
 import tensorflow as tf
 import torch
 
+
 def seed_everything(seed=1234):
     random.seed(seed)
     os.environ['PYTHONHASHSEED'] = str(seed)
     np.random.seed(seed)
     tf.keras.utils.set_random_seed(1)
     tf.config.experimental.enable_op_determinism()
-    
-    
+
+
 def torch_seed_everything(seed=1234):
     random.seed(seed)
     os.environ['PYTHONHASHSEED'] = str(seed)
@@ -19,4 +20,4 @@ def torch_seed_everything(seed=1234):
     torch.manual_seed(seed)
     torch.cuda.manual_seed(seed)
     torch.backends.cudnn.benchmark = True
-    torch.backends.cudnn.deterministic = True    
\ No newline at end of file
+    torch.backends.cudnn.deterministic = True
\ No newline at end of file
diff --git a/ptype/visualization_utils.py b/ptype/visualization_utils.py
index 5f4d762..6071348 100644
--- a/ptype/visualization_utils.py
+++ b/ptype/visualization_utils.py
@@ -16,9 +16,8 @@
 from cartopy import crs as ccrs
 from cartopy import feature as cfeature
 import imageio
-from PIL import Image
 from pathlib import Path
-from datetime import datetime, timedelta
+from datetime import datetime
 import xarray as xr
 from scipy.ndimage.filters import gaussian_filter
 
@@ -34,14 +33,17 @@
  "okla":[39.0, 31.0, -90.0, -106.0]}
  
 # colors = {0:'lime', 1:'darkturquoise', 2:'red', 3:'black'}
-colors = {0:'lime', 1:'dodgerblue', 2:'red', 3:'black'}
+colors = {0: 'lime', 1: 'dodgerblue', 2: 'red', 3: 'black'}
 datapath = "/glade/p/cisl/aiml/ai2es/winter_ptypes/precip_rap/"
 
-def ptype_map(datatype, starttime, endtime, gifname, imgsavepath="gif_images", gifsavepath="gifs", coords="na", duration=0.5):
+
+def ptype_map(datatype, starttime, endtime, gifname,
+              imgsavepath="gif_images", gifsavepath="gifs", coords="na",
+              duration=0.5):
     """
     Create and save GIF of P-Type data over specific CONUS region and time range.
-    
-    :param datatype: 
+
+    :param datatype:
     :param starttime:
     :param endtime:
     :param gifname:
@@ -55,7 +57,7 @@ def ptype_map(datatype, starttime, endtime, gifname, imgsavepath="gif_images", g
     enddate = datetime.strptime(endtime, "%Y%m%d %H:%M:%S")    
     time_range = pd.date_range(startdate, enddate, freq="h").strftime("%Y%m%d %H:%M:%S")
     coords = coord_dict[coords]
-    
+
     # Account for differences between mPING and ASOS.
     if datatype == "mping":
         if enddate >= datetime.strptime("20180101", "%Y%m%d"):