update to charts

app.py

@@ -7,6 +7,7 @@ from src.basic_plot import basic_chart
 from src.map_viz import calling_map_viz
 from src.data_ingestion import daily_average
 from src.heatmap import HeatMap
+from src.data_ingestion import remove_previous_view, merge_volumes
 
 def fetch_data():
     # comment out for local testing, but be sure to include after testing
@@ -21,8 +22,9 @@ def fetch_data():
     return counts_df
 
 def main():
-
     counts_df = fetch_data()
+    counts_df = remove_previous_view(counts_df)
+    counts_df = merge_volumes(counts_df)
     # st.set_page_config(layout="wide")
     height = 650
 
@@ -38,34 +40,48 @@ def main():
     checkbox_one = st.sidebar.checkbox('Overall Traffic', value = True) # rename as necessary
     checkbox_two = st.sidebar.checkbox('Traffic Map', value = True)
     checkbox_three = st.sidebar.checkbox('Heat Map', value = True)
+    view_options = list(counts_df["view"].unique())
+    view_options.append('All')
+    view = st.sidebar.selectbox("Choose View", options=view_options, index = view_options.index("Woodlands - to Johor"))
+    if view != 'All':
+        st.header(f"Showing Traffic for {view}")
+        counts_df = counts_df[counts_df['view'] == view]
 
     if checkbox_one:
-        st.plotly_chart(basic_chart(counts_df),use_container_width=True)
+        st.subheader("Overall Traffic")
+        plot = st.selectbox("Choose Plot", options=["Day", "Hour", "Raw"], index = 0)
+        st.plotly_chart(basic_chart(counts_df, plot = plot),use_container_width=True)
 
     if checkbox_two:
+        st.subheader("Traffic Map")
         st.pyplot(calling_map_viz(counts_df))
     
     if checkbox_three:
 
         heatmap = HeatMap(counts_df)
-
-        st.plotly_chart(heatmap.vehicle_count_bar())
+        
+        # st.header("Mean Vehicle Count by Day of Week")
+        # st.plotly_chart(heatmap.vehicle_count_bar())
+        st.subheader("Heatmap")
         st.plotly_chart(heatmap.heatmap())
-
-        hour_choice = st.selectbox(
-            "Choose Hour",
-            options=[
+        hours = [
                 "00:00", "01:00", "02:00", "03:00", "04:00", "05:00",
                 "06:00", "07:00", "08:00", "09:00", "10:00", "11:00",
                 "12:00", "13:00", "14:00", "15:00", "16:00", "17:00",
                 "18:00", "19:00", "20:00", "21:00", "22:00", "23:00",
-            ],
-            key = "hour"
+            ]
+        hour_choice = st.selectbox(
+            "Choose Hour",
+            options= hours,
+            key = "hour", index = hours.index("08:00")
         )
+        
+        st.subheader(f"Traffic Volume of Each Day at {hour_choice}")
         st.plotly_chart(heatmap.update_hour_bar_chart(hour_choice))
-
-        day_choice = st.selectbox("Choose Day of the Week", ["Monday", "Tuesday", "Wednesday", 
-                                                "Thursday", "Friday","Saturday", "Sunday"], key = "day")
+        days = ["Monday", "Tuesday", "Wednesday", 
+                "Thursday", "Friday","Saturday", "Sunday"]
+        day_choice = st.selectbox("Choose Day of the Week", options = days, key = "day", index = days.index("Saturday"))
+        st.subheader(f"Traffic Volume of Each Hour on {day_choice}")
         st.plotly_chart(heatmap.update_day_bar_chart(day_choice))
         
 

src/basic_plot.py

@@ -4,27 +4,70 @@ import plotly.express as px
 from datasets import load_dataset
 import os
 
-    
-def basic_chart(counts_df):
+def basic_chart(counts_df, plot, hovermode = False):
+
     # data processing
-    counts_df['traffic'] = counts_df['car'] + counts_df['motorcycle'] + counts_df['large_vehicle']
-    counts_df['datetime'] = pd.to_datetime(counts_df['date'] + ' ' + counts_df['time'])
-    counts_df['weekday'] = counts_df['datetime'].dt.strftime('%A')
-    counts_df['hour'] = counts_df['datetime'].dt.strftime('%H')
-    
-    # plot types
-    plot = st.selectbox('Choose Plot', options=['Day','Hour','Raw'])
-    
-    # view types
-    view = st.selectbox('Choose View', options=counts_df['view'].unique())
-    filtered_views = counts_df[counts_df['view'] == view]
-    
+    counts_df["traffic"] = (
+        counts_df["car"] + counts_df["motorcycle"] + counts_df["large_vehicle"]
+    )
+    counts_df["datetime"] = pd.to_datetime(counts_df["date"] + " " + counts_df["time"])
+    counts_df["weekday"] = counts_df["datetime"].dt.strftime("%A")
+    counts_df["hour"] = counts_df["datetime"].dt.strftime("%H")
+
+    # print (counts_df.head())
+
+    # get the mean by the weekday
+    date_view = counts_df.groupby(by=["view", "weekday"]).mean().round(1)
+    date_view = date_view.reset_index()
+
+    # get the mean by the time
+    time_view = counts_df.groupby(by=["view", "hour"]).mean().round(1)
+    time_view = time_view.reset_index()
+
     # conditional views
-    if plot == 'Day':    
-        fig = px.bar(filtered_views, x='weekday', y='traffic')
-    elif plot == 'Hour':
-        fig = px.bar(filtered_views, x='hour', y='traffic')
-    elif plot == 'Raw':
-        fig = px.bar(filtered_views, x='datetime', y='traffic')
-        
-    return fig
+    if plot == "Day":
+        # filtered_view_day = date_view[date_view["view"] == view]
+        fig = px.bar(
+            date_view,
+            x="weekday",
+            y="traffic",
+            category_orders = {'weekday': ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']},
+            labels={
+                "weekday": "Day",
+                "traffic": "Traffic",
+            },
+        )
+        fig.update_layout(
+            yaxis_visible=True, yaxis_showticklabels=False, hovermode=hovermode, xaxis_title = '', yaxis_title = ''
+        )
+    elif plot == "Hour":
+        # filterd_view_time = time_view[time_view["view"] == view]
+        fig = px.bar(
+            time_view,
+            x="hour",
+            y="traffic",
+            labels={
+                "hour": "Hour",
+                "traffic": "Traffic",
+            },
+        )
+        fig.update_layout(
+            yaxis_visible=True, yaxis_showticklabels=False, hovermode=hovermode
+        )
+        fig.update_xaxes(tickmode = 'linear', type = 'category')
+    elif plot == "Raw":
+        # filtered_views = counts_df[counts_df["view"] == view]
+        fig = px.bar(
+            counts_df,
+            x="datetime",
+            y="traffic",
+            labels={
+                "datetime": "Date and Time",
+                "traffic": "Traffic",
+            },
+        )
+        fig.update_layout(
+            yaxis_visible=True, yaxis_showticklabels=False, hovermode=hovermode
+        )
+
+    return fig

src/data_ingestion.py

@@ -1,6 +1,35 @@
 import pandas as pd
 import numpy as np
 
+def remove_previous_view(counts_df):
+    filtered_views_list = ['View_from_Second_Link_at_Tuas_to_sg',
+                    'View_from_Second_Link_at_Tuas_to_jh',
+                    'View_from_Tuas_Checkpoint_to_sg',
+                    'View_from_Tuas_Checkpoint_to_jh',
+                    'View_from_Woodlands_Causeway_Towards_Johor_to_sg',
+                    'View_from_Woodlands_Causeway_Towards_Johor_to_jh',
+                    'View_from_Woodlands_Checkpoint_Towards_BKE_to_sg',
+                    'View_from_Woodlands_Checkpoint_Towards_BKE_to_jh']
+    counts_df = counts_df[counts_df['view'].isin(filtered_views_list)]
+    return counts_df
+
+def merge_volumes(counts_df):
+    merge_groups = {"Tuas - to SG": ["View_from_Second_Link_at_Tuas_to_sg", "View_from_Tuas_Checkpoint_to_sg"],
+                "Tuas - to Johor": ['View_from_Second_Link_at_Tuas_to_jh', 'View_from_Tuas_Checkpoint_to_jh'],
+                "Woodlands - to SG": ['View_from_Woodlands_Causeway_Towards_Johor_to_sg', 'View_from_Woodlands_Checkpoint_Towards_BKE_to_sg'],
+                "Woodlands - to Johor": ['View_from_Woodlands_Causeway_Towards_Johor_to_jh', 'View_from_Woodlands_Checkpoint_Towards_BKE_to_jh']}
+    def apply_merge_groups(row):
+        for key, value in merge_groups.items():
+            if row in value:
+                return key
+    counts_df['merge_group'] = counts_df['view'].apply(apply_merge_groups)
+    counts_df = counts_df.groupby(by = ['merge_group', 'date', 'time']).sum(numeric_only = True)
+    counts_df = counts_df.reset_index()
+    counts_df = counts_df.rename(columns={"merge_group": "view"})
+
+    return counts_df
+    
+
 def daily_average(counts_df):
     
     filtered_views_list = ['View_from_Second_Link_at_Tuas_to_sg',

src/heatmap.py

@@ -7,6 +7,7 @@ import plotly.express as px
 from plotly.subplots import make_subplots
 import plotly.graph_objects as go
 import streamlit as st
+from src.data_ingestion import remove_previous_view, merge_volumes
 
 
 def clean_data(df):
@@ -22,21 +23,27 @@ def clean_data(df):
 
 class HeatMap:
     def __init__(self, counts_df):
+
+        # counts_df = counts_df[counts_df['view'] == view]
+        
         self.df = clean_data(counts_df)
+        
         new = (
             self.df.groupby(["hour", "day"])
-            .sum()
+            .mean(numeric_only = True)
             .drop(columns=["car", "large_vehicle"])
             .reset_index()
         )
+        
         table = pd.pivot_table(
             new, values="vehicle", index=["day"], columns=["hour"]
         ).reset_index()
-        self.table = table.reindex([1, 5, 6, 4, 0, 2, 3])
+        
+        self.table = table.reindex([1, 5, 6, 4, 0, 2, 3]).round(1)
 
 
     def vehicle_count_bar(self):
-        new_df = self.df.groupby(["day"]).sum().reset_index()
+        new_df = self.df.groupby(["day"]).mean(numeric_only = True).round(1).reset_index()
         new_df = new_df.reindex([1, 5, 6, 4, 0, 2, 3])
 
         veh_count_fig = px.bar(
@@ -47,6 +54,7 @@ class HeatMap:
             text_auto=True,
             labels={"day": "Day of the Week", "vehicle": "Vehicle Count"},
         )
+        veh_count_fig.update_layout(showlegend=False)
 
         return veh_count_fig
 
@@ -57,59 +65,71 @@ class HeatMap:
         hm_fig = px.imshow(
             new_table,
             labels=dict(
-                x="Hour of the Day", y="Day of the Week", color="Causeway Traffic"
+                x="Hour of the Day", y="Day of the Week", color="Traffic"
             ),
             x=[
-                "12am",
-                "1am",
-                "2am",
-                "3am",
-                "4am",
-                "5am",
-                "6am",
-                "7am",
-                "8am",
-                "9am",
-                "10am",
-                "11am",
-                "12pm",
-                "1pm",
-                "2pm",
-                "3pm",
-                "4pm",
-                "5pm",
-                "6pm",
-                "7pm",
-                "8pm",
-                "9pm",
-                "10pm",
-                "11pm",
+                "00",
+                "01",
+                "02",
+                "03",
+                "04",
+                "05",
+                "06",
+                "07",
+                "08",
+                "09",
+                "10",
+                "11",
+                "12",
+                "13",
+                "14",
+                "15",
+                "16",
+                "17",
+                "18",
+                "19",
+                "20",
+                "21",
+                "22",
+                "23",
             ],
             y=[
-                "Monday",
-                "Tuesday",
-                "Wednesday",
-                "Thursday",
-                "Friday",
-                "Saturday",
-                "Sunday",
+                "Mon",
+                "Tue",
+                "Wed",
+                "Thu",
+                "Fri",
+                "Sat",
+                "Sun",
             ],
-            text_auto=True,
+            # text_auto=True,
+        )
+        hm_fig.update_xaxes(side="top", tickmode = 'linear', type = 'category')
+        hm_fig.update_layout(yaxis_visible=True, 
+                             yaxis_showticklabels=True,
+                             yaxis_title = ''
         )
-        hm_fig.update_xaxes(side="top")
-
         return hm_fig
 
     def update_hour_bar_chart(self, hour="08:00"):
-
+        
         fig_hours = px.bar(
             self.table,
             x="day",
             y=str(hour),
             color="day",
-            text_auto=True,
+            # text_auto=True,
             labels={"day": "Day of the Week"},
         )
+        fig_hours.update_layout(showlegend=False)
+        fig_hours.update_layout(yaxis_visible=False, 
+                             yaxis_showticklabels=False,
+                             yaxis_title = False
+        )
+        fig_hours.update_layout(xaxis_visible=True, 
+                             xaxis_showticklabels=True,
+                             xaxis_title = ''
+        )
         return fig_hours
 
     def update_day_bar_chart(self, day="Saturday"):
@@ -127,13 +147,24 @@ class HeatMap:
         ]
         t = t.reset_index()
 
+        t['hour'] = t['hour'].str[:2]
+
         fig_days = px.bar(
             t,
             x="hour",
             y=str(day),
             color=str(day),
-            text_auto=True,
+            # text_auto=True,
             labels={"hour": "Count of Each Hour"},
         )
-
+        fig_days.update_layout(showlegend=False)
+        fig_days.update_layout(yaxis_visible=False, 
+                             yaxis_showticklabels=False,
+                             yaxis_title = False
+        )
+        fig_days.update_layout(xaxis_visible=True, 
+                             xaxis_showticklabels=True,
+                             xaxis_title = '',
+        )
+        fig_days.update_xaxes(tickangle=0, tickmode = 'linear', type = 'category')
         return fig_days

src/map_viz.py

@@ -10,6 +10,8 @@ import numpy as np
 import requests
 import pandas as pd
 
+import warnings
+warnings.filterwarnings("ignore")
 
 def calling_map_viz(counts_df):
     r = "svg/snazzy-image-01.svg"