Update app.py

app.py

@@ -140,9 +140,9 @@ def parse_weather_data(data):
 def calculate_total_new_snow(df):
     """
     Calculate total new snow by:
-    1. Looking at values before 9 AM each day
-    2. Ensuring reasonable hourly change rates (max 3 inches/hour)
-    3. Allowing high totals if they accumulate gradually
+    1. Summing 3-hour snowfall amounts within each day period
+    2. Using 9 AM as the daily reset point
+    3. Filtering out obvious anomalies (>9 inches in 3 hours)
     
     Parameters:
     df (pandas.DataFrame): DataFrame with datetime and snowfall_3hr columns
@@ -161,60 +161,17 @@ def calculate_total_new_snow(df):
         lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
     )
     
-    def validate_sequence(values, times):
-        """
-        Validate a sequence of snow readings by checking hourly change rates.
-        Returns the most reliable final value.
-        """
-        if len(values) <= 1:
-            return values[-1] if len(values) > 0 else 0.0
-        
-        # Calculate hour-to-hour changes
-        changes = np.diff(values)
-        
-        # If any change is too large (> 3 inches per hour), we need to find the reliable sequence
-        if any(change > 3.0 for change in changes):
-            # Find the longest sequence of reasonable changes
-            valid_indices = [0]  # Start with first reading
-            current_value = values[0]
-            
-            for i in range(1, len(values)):
-                change = values[i] - values[i-1]
-                if 0 <= change <= 3.0:  # Allow only positive changes up to 3 inches
-                    valid_indices.append(i)
-                    current_value = values[i]
-                else:
-                    # Check if this might be a reset (value lower than previous)
-                    if values[i] < values[i-1]:
-                        # If it looks like a reset, start new sequence
-                        current_value = values[i]
-                        valid_indices.append(i)
-                    else:
-                        # If it's a spike, ignore it
-                        values[i] = current_value
-            
-            # Return the last valid value
-            return values[valid_indices[-1]]
-        else:
-            # If all changes are reasonable, use the last value
-            return values[-1]
-    
     def process_daily_snow(group):
-        """Process snow measurements for a single day period"""
-        # Only look at measurements before 9 AM
-        morning_data = group[group['datetime'].dt.hour < 9].copy()
-        
-        if len(morning_data) == 0:
-            return 0.0
-        
+        """Sum up the 3-hour snowfall amounts for each day period"""
         # Sort by time to ensure proper sequence
-        morning_data = morning_data.sort_values('datetime')
+        group = group.sort_values('datetime')
         
-        values = morning_data['snowfall_3hr'].values
-        times = morning_data['datetime'].values
+        # Filter out obvious anomalies (more than 9 inches in 3 hours)
+        MAX_THREE_HOUR_SNOW = 9.0  # Maximum reasonable snow in 3 hours
+        valid_snow = group['snowfall_3hr'].where(group['snowfall_3hr'] <= MAX_THREE_HOUR_SNOW, 0.0)
         
-        # Validate the sequence and get the final value
-        daily_total = validate_sequence(values, times)
+        # Sum up all valid 3-hour amounts
+        daily_total = valid_snow.sum()
         
         return daily_total
     
@@ -225,107 +182,7 @@ def calculate_total_new_snow(df):
 
 def create_daily_snow_plot(df, ax):
     """
-    Create an improved daily snow plot focusing on pre-9 AM totals
-    with validation of hourly change rates
-    """
-    # Create a copy of the dataframe
-    snow_df = df[['datetime', 'snowfall_3hr']].copy()
-    
-    # Create day groups based on 9 AM reset
-    snow_df['day_group'] = snow_df['datetime'].apply(
-        lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
-    )
-    
-    # Calculate daily totals using the improved method
-    daily_snow = snow_df.groupby('day_group').apply(process_daily_snow).reset_index()
-    daily_snow.columns = ['date', 'new_snow']
-    
-    # Create the bar plot
-    ax.bar(daily_snow['date'], daily_snow['new_snow'], color='blue')
-    ax.set_title('Daily New Snow (9 AM Reset)', pad=20)
-    ax.set_xlabel('Date')
-    ax.set_ylabel('New Snow (inches)')
-    ax.tick_params(axis='x', rotation=45)
-    ax.grid(True, axis='y', linestyle='--', alpha=0.7)
-
-def create_daily_snow_plot(df, ax):
-    """
-    Create an improved daily snow plot that accounts for 9 AM reset
-    and filters out anomalous readings
-    
-    Parameters:
-    df (pandas.DataFrame): DataFrame with weather data
-    ax (matplotlib.axes.Axes): Axes object to plot on
-    """
-    # Create a copy of the dataframe
-    snow_df = df[['datetime', 'snowfall_3hr']].copy()
-    
-    # Create day groups based on 9 AM reset
-    snow_df['day_group'] = snow_df['datetime'].apply(
-        lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
-    )
-    
-    # Calculate daily totals using the improved method
-    daily_snow = snow_df.groupby('day_group').apply(process_daily_snow).reset_index()
-    daily_snow.columns = ['date', 'new_snow']
-    
-    # Create the bar plot
-    ax.bar(daily_snow['date'], daily_snow['new_snow'], color='blue')
-    ax.set_title('Daily New Snow (9 AM Reset)', pad=20)
-    ax.set_xlabel('Date')
-    ax.set_ylabel('New Snow (inches)')
-    ax.tick_params(axis='x', rotation=45)
-    
-    # Add grid for better readability
-    ax.grid(True, axis='y', linestyle='--', alpha=0.7)
-    
-    def process_daily_snow(group):
-        # Reset the index to work with the time series
-        group = group.sort_values('datetime').reset_index(drop=True)
-        
-        # Calculate hour-to-hour changes
-        group['snow_change'] = group['snowfall_3hr'].diff()
-        
-        # Mark values as anomalous if:
-        # 1. The hour-to-hour change is greater than 3 inches
-        # 2. The total value jumps by more than 3 inches from the previous reading
-        max_hourly_change = 3.0  # Maximum allowed change per hour
-        
-        is_anomaly = (
-            (abs(group['snow_change']) > max_hourly_change) | 
-            (group['snowfall_3hr'] > group['snowfall_3hr'].shift(1) + max_hourly_change)
-        )
-        
-        # Replace anomalous values with NaN
-        group.loc[is_anomaly, 'snowfall_3hr'] = np.nan
-        
-        # Forward fill NaN values with the last valid measurement
-        group['snowfall_3hr'] = group['snowfall_3hr'].fillna(method='ffill')
-        
-        # If no valid measurements exist, use backward fill
-        group['snowfall_3hr'] = group['snowfall_3hr'].fillna(method='bfill')
-        
-        # For each day, take the maximum valid value as the total new snow
-        if len(group) > 0:
-            return group['snowfall_3hr'].max()
-        return 0
-    
-    # Calculate daily snow totals
-    daily_totals = snow_df.groupby('day_group').apply(process_daily_snow)
-    
-    # Sum up all daily totals
-    total_snow = daily_totals.sum()
-    
-    return total_snow
-
-def create_daily_snow_plot(df, ax):
-    """
-    Create an improved daily snow plot that accounts for 9 AM reset
-    and filters out anomalous readings
-    
-    Parameters:
-    df (pandas.DataFrame): DataFrame with weather data
-    ax (matplotlib.axes.Axes): Axes object to plot on
+    Create a daily snow plot showing summed 3-hour amounts
     """
     # Create a copy of the dataframe
     snow_df = df[['datetime', 'snowfall_3hr']].copy()
@@ -335,116 +192,25 @@ def create_daily_snow_plot(df, ax):
         lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
     )
     
-    # Calculate daily totals using the improved method
+    # Calculate daily totals by summing 3-hour amounts
     daily_snow = snow_df.groupby('day_group').apply(process_daily_snow).reset_index()
     daily_snow.columns = ['date', 'new_snow']
     
     # Create the bar plot
     ax.bar(daily_snow['date'], daily_snow['new_snow'], color='blue')
-    ax.set_title('Daily New Snow (9 AM Reset)', pad=20)
+    ax.set_title('Daily New Snow (Sum of 3-hour amounts, 9 AM Reset)', pad=20)
     ax.set_xlabel('Date')
     ax.set_ylabel('New Snow (inches)')
     ax.tick_params(axis='x', rotation=45)
-    
-    # Add grid for better readability
     ax.grid(True, axis='y', linestyle='--', alpha=0.7)
     
-    def process_daily_snow(group):
-        # Reset the index to work with the time series
-        group = group.sort_values('datetime').reset_index(drop=True)
-        
-        # Calculate hour-to-hour changes
-        group['snow_change'] = group['snowfall_3hr'].diff()
-        
-        # Mark values as anomalous if:
-        # 1. The hour-to-hour change is greater than 3 inches
-        # 2. The total value jumps by more than 3 inches from the previous reading
-        max_hourly_change = 3.0  # Maximum allowed change per hour
-        
-        is_anomaly = (
-            (abs(group['snow_change']) > max_hourly_change) | 
-            (group['snowfall_3hr'] > group['snowfall_3hr'].shift(1) + max_hourly_change)
-        )
-        
-        # Replace anomalous values with NaN
-        group.loc[is_anomaly, 'snowfall_3hr'] = np.nan
-        
-        # Forward fill NaN values with the last valid measurement
-        group['snowfall_3hr'] = group['snowfall_3hr'].fillna(method='ffill')
-        
-        # If no valid measurements exist, use backward fill
-        group['snowfall_3hr'] = group['snowfall_3hr'].fillna(method='bfill')
-        
-        # For each day, take the maximum valid value as the total new snow
-        if len(group) > 0:
-            return group['snowfall_3hr'].max()
-        return 0
-    
-    # Calculate daily snow totals
-    daily_totals = snow_df.groupby('day_group').apply(process_daily_snow)
-    
-    # Sum up all daily totals
-    total_snow = daily_totals.sum()
-    
-    return total_snow
+    # Add value labels on top of each bar
+    for i, v in enumerate(daily_snow['new_snow']):
+        if v > 0:  # Only label bars with snow
+            ax.text(i, v, f'{v:.1f}"', 
+                   ha='center', va='bottom')
 
-def create_daily_snow_plot(df, ax):
-    """
-    Create an improved daily snow plot that accounts for 9 AM reset
-    and filters out anomalous readings
-    
-    Parameters:
-    df (pandas.DataFrame): DataFrame with weather data
-    ax (matplotlib.axes.Axes): Axes object to plot on
-    """
-    # Create a copy of the dataframe
-    snow_df = df[['datetime', 'snowfall_3hr']].copy()
-    
-    # Create day groups based on 9 AM reset
-    snow_df['day_group'] = snow_df['datetime'].apply(
-        lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
-    )
-    
-    # Calculate daily totals using the improved method
-    daily_snow = snow_df.groupby('day_group').apply(process_daily_snow).reset_index()
-    daily_snow.columns = ['date', 'new_snow']
-    
-    # Create the bar plot
-    ax.bar(daily_snow['date'], daily_snow['new_snow'], color='blue')
-    ax.set_title('Daily New Snow (9 AM Reset)', pad=20)
-    ax.set_xlabel('Date')
-    ax.set_ylabel('New Snow (inches)')
-    ax.tick_params(axis='x', rotation=45)
-    
-    # Add grid for better readability
-    ax.grid(True, axis='y', linestyle='--', alpha=0.7)
 
-def create_daily_snow_plot(df, ax):
-    """
-    Create an improved daily snow plot that accounts for 9 AM reset
-    
-    Parameters:
-    df (pandas.DataFrame): DataFrame with weather data
-    ax (matplotlib.axes.Axes): Axes object to plot on
-    """
-    # Create a copy of the dataframe
-    snow_df = df[['datetime', 'snowfall_3hr']].copy()
-    
-    # Create day groups based on 9 AM reset
-    snow_df['day_group'] = snow_df['datetime'].apply(
-        lambda x: x.date() if x.hour >= 9 else (x - pd.Timedelta(days=1)).date()
-    )
-    
-    # Calculate daily totals using the improved method
-    daily_snow = snow_df.groupby('day_group').apply(process_daily_snow).reset_index()
-    daily_snow.columns = ['date', 'new_snow']
-    
-    # Create the bar plot
-    ax.bar(daily_snow['date'], daily_snow['new_snow'], color='blue')
-    ax.set_title('Daily New Snow (9 AM Reset)', pad=20)
-    ax.set_xlabel('Date')
-    ax.set_ylabel('New Snow (inches)')
-    ax.tick_params(axis='x', rotation=45)
     
 def create_wind_rose(df, ax):
     """Create a wind rose plot"""