Update app.py

app.py

@@ -1,305 +1,194 @@
 import pandas as pd
-import pandas as pd
 import panel as pn
 import hvplot.pandas
-from itertools import cycle
-from bokeh.palettes import Reds9
-import folium
-raw_df = pd.read_csv('zomato_data.csv')
-zomato_df = raw_df.copy()
-rating_type_df = zomato_df['RATING_TYPE'].value_counts().reset_index()
-rating_type_df.rename(columns={'index':'RATING TYPE', 'RATING_TYPE':'COUNT OF RESTAURANTS'}, inplace=True)
-foodtruck_df = zomato_df[zomato_df['CUSINE TYPE'] == 'Food Truck']
-foodtruck_df.sort_values(by='RATING',ascending=False)
-
-
-# Read the CSV file into a DataFrame
-zomato_df = pd.read_csv('zomato_data.csv')
-
-# Count the occurrences of each cuisine type
-cuisine_counts = zomato_df['CUSINE TYPE'].value_counts()
-
-# Create the bar plot using hvplot
-bar_plot_cuisine = cuisine_counts.hvplot.bar(
-    color='#E10F14',
-    title='No. of Restaurants by Cuisine Type',
-    xlabel='Cuisine Type',
-    ylabel='Count',
-    width=900,
-    height=500
-).opts(xrotation=90)
-
-# Wrap the bar plot in a Panel object
-panel_cuisine = pn.panel(bar_plot_cuisine)
-
-# Create a DataFrame with the given data
-rating_type_df = pd.DataFrame({
-    'RATING TYPE': ['Average', 'Good', 'Very Good', 'Excellent', 'Poor', 'Very Poor'],
-    'COUNT OF RESTAURANTS': [4983, 4263, 1145, 96, 56, 4]
-})
-
-# Define the hvplot chart
-bar_plot_rating = rating_type_df.hvplot.bar(
-    x='RATING TYPE',
-    y='COUNT OF RESTAURANTS',
-    color='#E10F14',
-    title='Count of Restaurants by Rating Type',
-    xlabel='Rating Type',
-    ylabel='Count',
-    width=900,
-    height=500
-)
-
-# Wrap the bar plot in a Panel object
-panel_rating = pn.panel(bar_plot_rating)
-
-# Filter food trucks in Mumbai
-foodtruck_df = zomato_df[zomato_df['CUSINE TYPE'] == 'Food Truck']
-
-# Sort by rating in descending order and select the top result
-best_food_truck = foodtruck_df.sort_values(by='RATING', ascending=False).head()
+import numpy as np
+from math import radians, sin, cos, sqrt, asin
+uber_data = pd.read_csv(r'uber-raw-data-jul14.csv')
+type(uber_data.loc[0,'Date/Time'])
+uber_data['Date/Time'] = pd.to_datetime(uber_data['Date/Time'])
+uber_data['BinnedHour']=uber_data['Date/Time'].dt.floor('15min')
+uber_data['BinnedHour'].value_counts()
+DayMap={0:'Monday', 1:'Tuesday', 2:'Wednesday', 3:'Thursday', 4:'Friday', 5:'Saturday', 6:'Sunday'}
+uber_data['Day']=uber_data['BinnedHour'].dt.weekday.map(DayMap)
+uber_data['Date']=uber_data['BinnedHour'].dt.date
+uber_data['Day']=pd.Categorical(uber_data['Day'],categories=['Monday','Tuesday','Wednesday','Thursday','Friday','Saturday','Sunday'],ordered=True)
+uber_data['Time']=uber_data['BinnedHour'].dt.time
+weekly_data1 = uber_data.groupby(['Date','Day','Time']).count().dropna().rename(columns={'BinnedHour':'Rides'})['Rides'].reset_index()
+daywise = weekly_data1.groupby('Day').sum('Day')
+# Assuming you have the 'uber_data' DataFrame already defined
+
+# --- Code 1 ---
+# Calculate the value counts and sort by index
+value_counts = uber_data['BinnedHour'].dt.day.value_counts().sort_index()
+
+# Create a DataFrame from the value counts
+df = pd.DataFrame({'Days': value_counts.index, 'Rides': value_counts.values})
+
+# Create a Panel object for the Uber rides graph
+pn.extension('plotly')
+pn.config.sizing_mode = 'stretch_width'
+uber_rides_graph = df.hvplot.bar(x='Days', y='Rides', color='black', xlabel='Days', ylabel='Rides',
+                                 rot=0, title='Uber Rides per day in July 2014 at NYC',
+                                 height=400, width=800)
+
+# --- Code 2 ---
+# Calculate the value counts and sort by index
+value_counts = uber_data['BinnedHour'].value_counts().sort_index()
+
+# Create a DataFrame from the value counts
+df = pd.DataFrame({'BinnedHour': value_counts.index, 'Rides': value_counts.values})
+
+# Create a Bokeh figure for the interactive DataFrame graph
+interactive_df_figure = df.hvplot.line(x='BinnedHour', y='Rides', color='black', alpha=0.8,
+                                       title='Uber Rides every 15 mins in the month of July at NYC',
+                                       xlabel='Days', ylabel='No. of Rides',
+                                       height=400, width=800)
+
+# Create a Panel object with the Bokeh figure
+interactive_df_pane = pn.pane.HoloViews(interactive_df_figure)
+
+# --- Code 3 ---
+# Extracting day of the week from the 'BinnedHour' column
+uber_data['BinnedHour'] = pd.to_datetime(uber_data['BinnedHour'])
+uber_data['BinnedHour'].value_counts()
+DayMap = {0: 'Monday', 1: 'Tuesday', 2: 'Wednesday', 3: 'Thursday', 4: 'Friday', 5: 'Saturday', 6: 'Sunday'}
+uber_data['Day'] = uber_data['BinnedHour'].dt.weekday.map(DayMap)
+uber_data['Date'] = uber_data['BinnedHour'].dt.date
+uber_data['Day'] = pd.Categorical(uber_data['Day'],
+                                  categories=['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday',
+                                              'Sunday'],
+                                  ordered=True)
+uber_data['Time'] = uber_data['BinnedHour'].dt.time
+
+# Grouping by Date, Day, and Time to get the count of rides for each time slot
+weekly_data = uber_data.groupby(['Date', 'Day', 'Time']).count().dropna().rename(columns={'BinnedHour': 'Rides'})[
+    'Rides'].reset_index()
+
+# Summing up the rides per day
+daywise = weekly_data.groupby('Day')['Rides'].sum()
+df_total_rides = pd.DataFrame({'Days': daywise.index, 'Rides': daywise.values})
+
+# Create a Panel object for the 'Total Rides per Day' graph
+total_rides_graph = df_total_rides.hvplot.bar(x='Days', y='Rides', color='black', xlabel='Days', ylabel='Total Rides',
+                                             rot=0, title='Total Rides per Day',
+                                             height=400, width=800,
+                                             value_label=True)  # Display total value when hovering
+
+# --- Code 4 ---
+# Your original data processing
+weekly_data = weekly_data.groupby(['Day', 'Time']).mean('Rides')
+weekly_data1 = weekly_data.unstack(level=0)
+
+# Create a Panel object
+avg_rides_graph = pn.panel(weekly_data1.T.mean().hvplot(c='black', xlabel='Date', ylabel='Average rides',
+                                                        xticks=10, title='Average Uber rides on any day in July 2014 at NYC',
+                                                        height=400, width=800))
+
+# --- Code 5 ---
+# Countplot using hvplot
+BaseMapper = {'B02512': 'Unter', 'B02598': 'Hinter', 'B02617': 'Weiter', 'B02682': 'Schmecken', 'B02764': 'Danach-NY'}
+plot_top_rides_city = uber_data['Base'].map(BaseMapper).value_counts().hvplot(kind='bar', rot=0, xlabel='Base', ylabel='Total rides', color='black',
+                                                                             title='CountPlot: Total uber rides vs Base - July 2014, NYC', height=400, width=800)
+
+# --- Code 6 ---
+# Your code 6 as provided
+metro_art_coordinates = (40.7794, -73.9632)
+empire_state_building_coordinates = (40.7484, -73.9857)
+
+def haversine(coordinates1, coordinates2):
+    lat1, lon1 = coordinates1
+    lat2, lon2 = coordinates2
+    
+    # Convert to radians and apply Haversine formula
+    lon1, lat1, lon2, lat2 = map(radians, [lon1, lat1, lon2, lat2])
+    dlon = lon2 - lon1
+    dlat = lat2 - lat1
+    
+    a = sin(dlat/2)**2 + cos(lat1)*cos(lat2)*sin(dlon/2)**2
+    c = 2 * asin(sqrt(a))
+    r = 3956
+    return c * r
 
-# Create the bar plot using hvplot
-bar_plot_best_food_truck = best_food_truck.hvplot.bar(
-    x='NAME',
-    y='PRICE',
-    color='#E10F14',
-    title='Best Food Truck in Mumbai: Price vs. Name',
-    xlabel='Food Truck Name',
-    ylabel='Price',
-    hover_cols=['RATING', 'REGION', 'CUSINE_CATEGORY'],
-    rot=90,
-    width=900,
-    height=500
-)
+# Assuming `uber_data` is a DataFrame containing 'Lat' and 'Lon' columns
+# Calculate distances from 'metro_art_coordinates' and 'empire_state_building_coordinates'
+uber_data['Distance MM'] = uber_data[['Lat', 'Lon']].apply(lambda x: haversine(metro_art_coordinates, tuple(x)), axis=1)
+uber_data['Distance ESB'] = uber_data[['Lat', 'Lon']].apply(lambda x: haversine(empire_state_building_coordinates, tuple(x)), axis=1)
 
-# Wrap the bar plot in a Panel object
-panel_best_food_truck = pn.panel(bar_plot_best_food_truck)
+# Count the number of rides within 0.25 miles of each location
+# print((uber_data[['Distance MM', 'Distance ESB']] < 0.25).sum())
 
-# Filter seafood restaurants in Mumbai
-seafood_df = zomato_df[zomato_df['CUSINE_CATEGORY'].notna() & zomato_df['CUSINE_CATEGORY'].str.contains('Seafood')]
+# Create distance range and count the number of rides within each distance
+distance_range = np.arange(0.1, 5.1, 0.1)
+distance_data = [(uber_data[['Distance MM', 'Distance ESB']] < dist).sum() for dist in distance_range]
+distance_data = pd.concat(distance_data, axis=1)
+distance_data = distance_data.T
+distance_data.index = distance_range
+distance_data = distance_data.rename(columns={'Distance MM': 'CloserToMM', 'Distance ESB': 'CloserToESB'})
 
-# Get top 10 seafood restaurants in Mumbai, sorted by rating
-top_seafood_df = seafood_df.sort_values(by='RATING', ascending=False).head(10)
+pn.extension('bokeh')
 
-# Create the bar plot using hvplot
-bar_plot_top_seafood = top_seafood_df.hvplot.bar(
-    x='NAME',
-    y='PRICE',
-    color='#E10F14',
-    title='Top 10 Seafood Restaurants in Mumbai: Price vs. Name',
-    xlabel='Restaurant Name',
-    ylabel='Price',
-    hover_cols=['RATING', 'REGION', 'CUSINE_CATEGORY'],
-    rot=90,
-    width=900,
-    height=500
-)
+# Create the hvplot figure with customized colors
+fig = distance_data.hvplot(height=400, width=800, color=['black', 'grey']).opts(title='Number of Rides Closer to ESB and MM',
+                                                                            xlabel='Threshold Radius(mi)',
+                                                                            ylabel='Rides')
 
-# Wrap the bar plot in a Panel object
-panel_top_seafood = pn.panel(bar_plot_top_seafood)
+# Create a panel with the figure
+fig_panel = pn.panel(fig)
 
 # Define Panel widgets
 yaxis_radio = pn.widgets.RadioButtonGroup(
     name='Y axis',
-    options=['Cuisine Type', 'Rating Type', 'Best Food Truck', 'Top 10 Seafood', 'Highest Rated', 'Top Avg Price', 'Chinese Resto', 'Price vs Rating', 'Region vs Price', 'Map'],
-    button_type='danger',
-    inline=True,
-    value='Cuisine Type'
+    options=['Rides vs Days', '15 min of Uber', 'Total Rides per Day', 'Avg Rides per Day', 'Top Rides City', 'Predicting Distance'],
+    button_type='light',
+    button_style='solid',
+    inline=True
 )
 
 # Define the Panel layout
 panel_layout = pn.Column(
-    pn.Row(yaxis_radio)
+    yaxis_radio,
+    pn.pane.HoloViews(uber_rides_graph),
 )
 
-# Create the map centered at Mumbai with dark mode
-mumbai_map = folium.Map(location=[19.0760, 72.8777], zoom_start=12, tiles="StamenTonerBackground")
-
-# Add a marker for Mumbai
-folium.Marker(
-    location=[19.0760, 72.8777],
-    popup='<b>Mumbai</b>',
-    icon=folium.Icon(color='red', icon_color='white', icon='heart', prefix='fa')
-).add_to(mumbai_map)
-
-# Add markers for the specified locations
-locations = [
-    {'name': 'Hitchki', 'region': 'Bandra', 'rating': '4.8', 'latitude': 19.0590, 'longitude': 72.8292, 'cuisine': 'Indian'},
-    {'name': 'Downtown China', 'region': 'Andheri', 'rating': '4.9', 'latitude': 19.1136, 'longitude': 72.8697, 'cuisine': 'Chinese'},
-    {'name': 'The Northern Vibe', 'region': 'Powai', 'rating': '4.7', 'latitude': 19.1187, 'longitude': 72.9073, 'cuisine': 'Continental'},
-    {'name': 'Rajdhani', 'region': 'Ghatkopar', 'rating': '4.8', 'latitude': 19.0866, 'longitude': 72.9081, 'cuisine': 'Indian'},
-    {'name': 'Trumpet Sky Lounge', 'region': 'Andheri', 'rating': '4.9', 'latitude': 19.1189, 'longitude': 72.8537, 'cuisine': 'International'},
-    {'name': 'Dessertino', 'region': 'Kandivali', 'rating': '4.7', 'latitude': 19.2128, 'longitude': 72.8376, 'cuisine': 'Desserts'}
-]
-
-for location in locations:
-    popup_content = f"<b>Name:</b> {location['name']}<br><b>Region:</b> {location['region']}<br><b>Rating:</b> {location['rating']}<br><b>Cuisine:</b> {location['cuisine']}"
-    if location['name'] == 'Dessertino':
-        icon = folium.Icon(color='red', icon_color='white', icon='coffee', prefix='fa')
-    else:
-        icon = folium.Icon(color='red', icon_color='white', icon='cutlery', prefix='fa')
-    folium.Marker(
-        location=[location['latitude'], location['longitude']],
-        popup=popup_content,
-        icon=icon
-    ).add_to(mumbai_map)
-    
-title_html = """
-<div style="font-size: 17px; font-weight: bold; text-align: left;">The best Restaurant to order food with best price and Quality</div>
-"""
-# Wrap the map in a Panel object
-panel_map = pn.pane.HTML(title_html + mumbai_map._repr_html_(), width=800, height=600)
-
 # Define the callback function for the radio button
 def update_chart(event):
-    if event.new == 'Cuisine Type':
-        panel_layout[1:] = [panel_cuisine]
-    elif event.new == 'Rating Type':
-        panel_layout[1:]= [panel_rating]
-    elif event.new == 'Best Food Truck':
-        panel_layout[1:] = [panel_best_food_truck]
-    elif event.new == 'Top 10 Seafood':
-        panel_layout[1:] = [panel_top_seafood]
-    elif event.new == 'Highest Rated':
-        # Filter the DataFrame for highest rated restaurants
-        highest_rated = zomato_df[zomato_df['RATING'] >= 4.7]
-
-        # Create the bar plot using hvplot
-        bar_plot_highest_rated = highest_rated.hvplot.bar(
-            x='NAME',
-            y='PRICE',
-            color='#E10F14',
-            title='Highest Rated Restaurants in Mumbai: Price vs. Name',
-            xlabel='Restaurant Name',
-            ylabel='Price',
-            hover_cols=['RATING', 'REGION', 'CUSINE_CATEGORY'],
-            rot=90,
-            width=900,
-            height=500
-        )
-
-        # Wrap the bar plot in a Panel object
-        panel_highest_rated = pn.panel(bar_plot_highest_rated)
-        panel_layout[1:] = [panel_highest_rated]
-    elif event.new == 'Top Avg Price':
-        # Filter the DataFrame for ratings greater than or equal to 4.5
-        filtered_df = zomato_df[zomato_df['RATING'] >= 4.5]
-
-        # Calculate the mean price for each combination of 'REGION' and 'CUSINE TYPE'
-        highest_rated_price_df = filtered_df.groupby(['REGION', 'CUSINE TYPE'])['PRICE'].mean().reset_index()
-
-        # Sort the DataFrame by 'REGION' in alphabetical order
-        highest_rated_price_df = highest_rated_price_df.sort_values('REGION')
-
-        # Create a scatter plot with rotated labels and star marker
-        scatter_plot_top_avg_price = highest_rated_price_df.hvplot.scatter(
-            x='REGION',
-            y='PRICE',
-            c='CUSINE TYPE',
-            cmap='Category10',
-            title='Avg Price Distribution of High-rated restaurants for each Cuisine Type',
-            size=100,  # Increase the marker size
-            rot=90,
-            width=900,
-            height=500,
-            marker='*',
-        )
-
-        # Create a Panel object with the scatter plot
-        panel_top_avg_price = pn.panel(scatter_plot_top_avg_price)
-        panel_layout[1:] = [panel_top_avg_price]
-    elif event.new == 'Chinese Resto':
-        zomato_df_cleaned = zomato_df.dropna(subset=['CUSINE_CATEGORY'])
-        chinese_df = zomato_df_cleaned[zomato_df_cleaned['CUSINE_CATEGORY'].str.contains('Chinese')]
-        chinese_rest_df = chinese_df.groupby(by='REGION').agg({'NAME': 'count', 'PRICE': 'mean'}).rename(columns={'NAME': 'COUNT OF RESTAURANTS'}).reset_index()
-        chinese_rest_df = chinese_rest_df.sort_values('COUNT OF RESTAURANTS', ascending=False).head(25)
-        bar_plot = chinese_rest_df.hvplot.bar(
-            x='REGION',
-            y='COUNT OF RESTAURANTS',
-            color='#E10F14',  # Set the color to red
-            title='No. of Chinese Restaurants by Places',
-            xlabel='Region',
-            ylabel='Count of Restaurants',
-            rot=90,
-            height=500,
-            width=900
-        )
-        layout = pn.Column(bar_plot)
-        panel_layout[1:] = [bar_plot]
-    elif event.new == 'Price vs Rating':
-        # Calculate the mean price and rating for each cuisine type
-        price_rating_df = zomato_df.groupby(['CUSINE TYPE', 'RATING'])['PRICE'].mean().reset_index()
-        hvplot_price_rating = price_rating_df.hvplot.line(
-            x='RATING',
-            y='PRICE',
-            by='CUSINE TYPE',
-            title='Price vs Rating by Cuisine Type',
-            xlabel='Rating',
-            ylabel='Price',
-            width=900,
-            height=500,
-            legend='bottom'  # Set the position of the legend to 'bottom'
-        )
-
-        # Set the number of legend columns
-        hvplot_price_rating.opts(legend_cols=6)  # Adjust the value to your desired maximum number of legend items per row
-
-        # Wrap the Hvplot plot in a Panel object
-        panel_price_vs_rating = pn.panel(hvplot_price_rating)
-        panel_layout[1:] = [panel_price_vs_rating]
-    elif event.new == 'Region vs Price':
-        region_price_df = zomato_df.groupby(['REGION'])['PRICE'].mean().reset_index()
-        scatter_plot = region_price_df.hvplot.scatter(
-            x='REGION',
-            y='PRICE',
-            cmap='Category10',
-            title='Relation between Region and Price',
-            size=100,  # Increase the marker size
-            rot=90,
-            width=900,
-            height=600,
-            marker='*',
-            color='red'
-        )
-        panel_region_vs_price = pn.Column(scatter_plot)
-        panel_layout[1:] = [panel_region_vs_price]
-    elif event.new == 'Map':
-        panel_layout[1:] = [panel_map]
+    if event.new == 'Rides vs Days':
+        panel_layout[1] = pn.pane.HoloViews(uber_rides_graph)
+    elif event.new == '15 min of Uber':
+        panel_layout[1] = interactive_df_pane
+    elif event.new == 'Total Rides per Day':
+        panel_layout[1] = total_rides_graph
+    elif event.new == 'Avg Rides per Day':
+        panel_layout[1] = avg_rides_graph
+    elif event.new == 'Top Rides City':
+        panel_layout[1] = plot_top_rides_city
+    elif event.new == 'Predicting Distance':
+        panel_layout[1] = fig_panel
 
 yaxis_radio.param.watch(update_chart, 'value')
-
-# Display the initial chart
-panel_layout.append(panel_cuisine)
+panel_layout.append
 
 # Display the Panel layout
 panel_layout
-dashboard = panel_layout
 import panel as pn
 pn.extension()  # Add this line to load the Panel extension
 
 # Layout using Template
 template = pn.template.FastListTemplate(
-    title='Zomato Mumbai Dashboard', 
+    title='Uber Analysis Dashboard', 
     sidebar=[
-        pn.pane.PNG('zomato.png', sizing_mode='scale_both'),
-        pn.pane.Markdown("# Performing Exploratory Data Analysis"),  
-        pn.pane.Markdown("1. How many restaurants are in Mumbai for each type of cuisine?"),
-        pn.pane.Markdown("2. What are the percentage of restaurants by Rating Type in Mumbai?"),
-        pn.pane.Markdown("3. Which are the Top 10 highest rated Seafood Restaurant in Mumbai?"),
-        pn.pane.Markdown("4. Which is the best Food Truck in Mumbai?"),
-        pn.pane.Markdown("5. Which places have the highest rated restaurant for each Cuisine Type in Mumbai?"),
-        pn.pane.Markdown("6. What is the Avg Price Distibution of highest rated restaurant for each Cuisine Type in Mumbai?"),
-        pn.pane.Markdown("7. Which areas have a large number of Chinese Restaurant Market?"),
-        pn.pane.Markdown("8. Is there a relation between Price and Rating by each Cuisine Type?"),
-        pn.pane.Markdown("9. Is there a relation between Region and Price?"),
-        pn.pane.Markdown("10. Can we map the best restraunt with high quality food?"),
-    ],
-    main = [pn.Row(pn.Column(dashboard)),
-            pn.Row(pn.pane.Markdown("Designed and Developed with ❤️ by Chitranshu Nagdawane © 2023"))
-           ],
-    accent_base_color="#E10F14",
-    header_background="#E10F14"
+        pn.pane.PNG('Uber2.png', sizing_mode='scale_both'),
+        pn.pane.Markdown("# Key Performance Indicators (KPIs) of the EDA"),  
+        pn.pane.Markdown("1. Let us visualize the total uber rides per day in the month of July 2014"),
+        pn.pane.Markdown("2. Let us have a more closer look at it, say every 15 minutes from July 1 to July 31."),
+        pn.pane.Markdown("3. Grouping weekly_data by days to plot total rides per week in july 2014."),
+        pn.pane.Markdown("4. Finding average rides on any day."),
+        pn.pane.Markdown("5. Now, let's try visualizing the relationship between Base and total number of rides in July 2014"),
+        pn.pane.Markdown("6. The number of rides predicted to Metropolitan Museum (MM) and Empire State Building (ESB)")],
+    main = [pn.Row(pn.Column(panel_layout))],
+    accent_base_color="#000000",
+    header_background="#000000"
 )
 
 template.servable()
+