Update app.py

app.py

@@ -2,10 +2,8 @@ import streamlit as st
 import pandas as pd
 import altair as alt
 
-# Set page title
-st.title(' UFO Data Visualization Analysis Report')
+st.title('UFO Data Visualization Analysis Report')
 
-# Load and process data
 @st.cache_data
 def load_data():
     columns = [
@@ -19,7 +17,6 @@ def load_data():
         header=None,
         names=columns
     )
-    # print(df)
     
     df['datetime'] = pd.to_datetime(df['datetime'])
     df['year'] = df['datetime'].dt.year
@@ -27,79 +24,136 @@ def load_data():
     df['shape'] = df['shape'].fillna('unknown').str.lower()
     return df
 
-
 df = load_data()
 
-# First visualization
 st.markdown("## 1. Temporal Trends in UFO Sightings")
 
+min_year = int(df['year'].min())
+max_year = int(df['year'].max())
+year_range = st.slider(
+    "Select Year Range",
+    min_value=min_year,
+    max_value=max_year,
+    value=(min_year, max_year)
+)
+
+time_agg = st.selectbox(
+    "Select Time Aggregation",
+    ['Yearly', 'Monthly']
+)
 
-yearly_counts = df.groupby('year').size().reset_index(name='count')
-yearly_viz = alt.Chart(yearly_counts).mark_line(
-    point=True
-).encode(
-    x=alt.X('year:Q', title='Year'),
+if time_agg == 'Yearly':
+    time_counts = df[df['year'].between(year_range[0], year_range[1])]\
+        .groupby('year').size().reset_index(name='count')
+    x_encoding = alt.X('year:Q', title='Year')
+else:
+    df['date'] = pd.to_datetime(df[['year', 'month']].assign(day=1))
+    time_counts = df[df['year'].between(year_range[0], year_range[1])]\
+        .groupby('date').size().reset_index(name='count')
+    x_encoding = alt.X('date:T', title='Date')
+
+base_chart = alt.Chart(time_counts).encode(
+    x=x_encoding,
     y=alt.Y('count:Q', title='Number of Sightings'),
-    tooltip=['year:Q', 'count:Q']
+    tooltip=['count:Q']
+)
+
+temporal_viz = (
+    base_chart.mark_area(opacity=0.3) +
+    base_chart.mark_line(color='steelblue') +
+    base_chart.mark_point(color='steelblue')
 ).properties(
     width=700,
     height=400
-)
+).interactive()
 
-st.altair_chart(yearly_viz)
+st.altair_chart(temporal_viz)
 
 st.markdown("""
 **(1) Features Highlighted**
 
-This visualization emphasizes the temporal evolution of UFO sightings from 1949 to 2013, highlighting both the overall trend and year-specific fluctuations in reporting frequency. The visualization reveals distinct patterns of increased reporting over time, with notable spikes in certain periods that could correlate with significant historical events or changes in reporting methods.
+This visualization emphasizes the temporal evolution of UFO sightings, with enhanced interactive features:
+- Adjustable time range using the slider
+- Choice between yearly and monthly aggregation
+- Interactive area and line combination chart
+- Hover tooltips for detailed information
 
 **(2) Design Choices**
 
 I implemented several key design elements for optimal data representation: 
-+ A line chart with point markers was chosen for its effectiveness in showing continuous time-series data while maintaining precise year-specific values
-+ Interactive tooltips were added to provide exact sighting counts
++ Combined area and line chart to show both trends and volume
++ Interactive time range selection for focused analysis
++ Multiple temporal granularity options
++ Synchronized area, line, and point markers for clear data representation
 
 **(3) Potential Improvements**
 
-Given more time, I would implement dual y-axes to show both sighting frequency and duration, and add the capability to filter by time periods and incorporate monthly/seasonal analysis options.
+Future enhancements could include seasonal analysis and correlation with historical events.
 """)
 
-# Second visualization
 st.markdown("## 2. Analysis of UFO Shape Distribution")
 
-shape_counts = df['shape'].value_counts().reset_index()
-shape_counts.columns = ['shape', 'count']
+all_shapes = sorted(df['shape'].unique())
+selected_shapes = st.multiselect(
+    "Select UFO Shapes to Display",
+    options=all_shapes,
+    default=all_shapes[:10]
+)
+
+sort_option = st.selectbox(
+    "Sort By",
+    ['Frequency', 'Alphabetical', 'Average Duration']
+)
+
+filtered_df = df[df['shape'].isin(selected_shapes)]
+
+if sort_option == 'Frequency':
+    shape_counts = filtered_df['shape'].value_counts().reset_index()
+    shape_counts.columns = ['shape', 'count']
+    sort_field = '-x'
+elif sort_option == 'Alphabetical':
+    shape_counts = filtered_df['shape'].value_counts().reset_index()
+    shape_counts.columns = ['shape', 'count']
+    sort_field = 'shape'
+else:
+    shape_counts = filtered_df.groupby('shape')['duration_seconds'].mean().reset_index()
+    shape_counts.columns = ['shape', 'count']
+    sort_field = '-count'
+
 shape_viz = alt.Chart(shape_counts).mark_bar().encode(
     y=alt.Y('shape:N', 
-            sort='-x',
+            sort=sort_field,
             title='UFO Shape'),
     x=alt.X('count:Q',
-            title='Number of Reports'),
-    color=alt.Color('count:Q',
-                    legend=None),
+            title='Number of Reports' if sort_option == 'Frequency' else 'Average Duration (seconds)'),
+    color=alt.Color('count:Q', scale=alt.Scale(scheme='viridis')),
     tooltip=['shape:N', 'count:Q']
 ).properties(
     width=700,
-    height=400
-)
+    height=max(len(selected_shapes) * 25, 400)
+).interactive()
 
 st.altair_chart(shape_viz)
 
-
 st.markdown("""
 **(1) Features Highlighted**
 
-This visualization focuses on the distribution of reported UFO shapes. The data shows clear preferences in how witnesses describe UFO shapes, with certain forms being consistently more common across reports. This analysis helps identify patterns in how people perceive and describe unidentified flying objects.
+This enhanced visualization now includes:
+- Multiple shape selection capability
+- Dynamic sorting options
+- Color encoding for better pattern recognition
+- Interactive tooltips and filtering
 
 **(2) Design Choices** 
 
-The visualization employs several intentional design elements: 
-+ A horizontal bar chart format was chosen to accommodate long shape labels and enable easy comparison of quantities
-+ Bars are sorted in descending order to immediately highlight the most common shapes
-+ The chart focuses on the top 10 shapes to maintain clarity and prevent information overload
-+ Interactive tooltips provide precise counts for each shape category.
+The visualization employs several interactive elements: 
++ Multi-select dropdown for shape filtering
++ Dynamic sorting options for different analytical perspectives
++ Color gradient to emphasize differences
++ Responsive height adjustment based on selected shapes
++ Interactive tooltips for detailed information
 
 **(3) Potential Improvements**
 
-With additional time, I would expand this visualization by adding temporal analysis to show how shape distributions have changed over decades, incorporate geographical analysis to reveal regional patterns in shape reporting.
+Future enhancements could include geographical distribution analysis and temporal trend analysis by shape.
 """)