Create app.py

app.py

@@ -0,0 +1,83 @@
+# Importing necessary libraries
+import pandas as pd
+import seaborn as sns
+import matplotlib.pyplot as plt
+from sklearn.tree import DecisionTreeRegressor
+from sklearn.linear_model import LinearRegression
+
+# Ensure seaborn is set up correctly
+sns.set(color_codes=True)
+
+# Load the dataset
+df = pd.read_csv('US_Accidents_March23.csv')
+
+# Convert 'Start_Time' to datetime format
+df['Start_Time'] = pd.to_datetime(df['Start_Time'], format='mixed', errors='coerce')
+
+# Analysis of hourly accidents distribution
+def plot_hourly_accidents(df):
+    # Create a 4x2 subplot grid
+    fig, axes = plt.subplots(4, 2, figsize=(18, 10))
+    plt.subplots_adjust(hspace=0.5)
+
+    # Gradient blue color palette
+    blue_palette = sns.light_palette("blue", n_colors=8, reverse=True)
+
+    for i in range(8):
+        ax = axes[i//2, i%2]
+        if i == 0:
+            sns.histplot(df['Start_Time'].dt.hour, bins=24, ax=ax, color=blue_palette[i])
+            ax.set_title("Overall Hourly Accident Distribution")
+        else:
+            day_data = df[df['Start_Time'].dt.dayofweek == i-1]
+            sns.histplot(day_data['Start_Time'].dt.hour, bins=24, ax=ax, color=blue_palette[i])
+            ax.set_title(f"Hourly Distribution: {day_data['Start_Time'].dt.day_name().iloc[0]}")
+        ax.set_xlabel("Hour")
+        ax.set_ylabel("Accidents")
+
+    plt.tight_layout()
+    plt.show()
+
+plot_hourly_accidents(df)
+
+# Analysis of weather conditions
+def plot_weather_conditions(df):
+    weather = df['Weather_Condition'].value_counts().head(15)
+    plt.figure(figsize=(30, 10))
+    sns.barplot(x=weather.index, y=weather.values, palette="Reds_r")
+    plt.xticks(rotation=45, fontsize=15)
+    plt.yticks(fontsize=15)
+    plt.xlabel("Weather Condition", fontsize=20)
+    plt.ylabel("Count", fontsize=20)
+    plt.title("Weather Condition vs Accidents", fontsize=30)
+    plt.show()
+
+plot_weather_conditions(df)
+
+# Additional Plots (as per your original code)
+# ... (include other plot functions as needed)
+
+# Prophet Model for Accident Prediction
+def prophet_model(df):
+    from prophet import Prophet
+
+    # Resampling data to get yearly count
+    df_yearly = df.resample('Y', on='Start_Time').size().reset_index(name='counts')
+    df_prophet = df_yearly.rename(columns={'Start_Time': 'ds', 'counts': 'y'})
+
+    model = Prophet()
+    model.fit(df_prophet)
+
+    future = model.make_future_dataframe(periods=5, freq='Y')
+    forecast = model.predict(future)
+
+    fig = model.plot(forecast)
+    plt.xlabel("Year")
+    plt.ylabel("Accidents")
+    plt.show()
+
+# Run the Prophet model function only if the Prophet package is installed
+try:
+    prophet_model(df)
+except ImportError:
+    print("Prophet package is not installed. Skipping the Prophet model prediction.")