Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,3 +1,105 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
import pandas as pd
|
| 3 |
import altair as alt
|
|
@@ -6,95 +108,73 @@ import altair as alt
|
|
| 6 |
url = 'https://github.com/UIUC-iSchool-DataViz/is445_data/raw/main/licenses_fall2022.csv'
|
| 7 |
df = pd.read_csv(url)
|
| 8 |
|
| 9 |
-
#
|
| 10 |
-
|
| 11 |
-
df[col] = df[col].astype(str)
|
| 12 |
-
|
| 13 |
-
# Convert integer columns to float
|
| 14 |
-
for col in df.select_dtypes(include=['int64', 'float64']).columns:
|
| 15 |
-
df[col] = pd.to_numeric(df[col], errors='coerce')
|
| 16 |
-
|
| 17 |
-
# Checking for missing values and handling them
|
| 18 |
-
df['Middle'] = df['Middle'].fillna('N/A') # Update Middle column without inplace
|
| 19 |
|
| 20 |
-
#
|
| 21 |
-
df
|
|
|
|
| 22 |
|
| 23 |
-
#
|
| 24 |
-
df
|
| 25 |
-
|
| 26 |
-
|
|
|
|
| 27 |
|
| 28 |
-
#
|
| 29 |
-
date_columns = ['Original Issue Date', 'Effective Date', 'Expiration Date', 'LastModifiedDate']
|
| 30 |
-
for col in date_columns:
|
| 31 |
-
df[col] = pd.to_datetime(df[col], errors='coerce', format='%m/%d/%Y')
|
| 32 |
-
|
| 33 |
-
# Add year columns
|
| 34 |
df['Original Issue Year'] = df['Original Issue Date'].dt.year
|
| 35 |
df['Effective Date Year'] = df['Effective Date'].dt.year
|
| 36 |
df['Expiration Date Year'] = df['Expiration Date'].dt.year
|
| 37 |
df['Last Modified Date Year'] = df['LastModifiedDate'].dt.year
|
| 38 |
|
| 39 |
-
#
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
# # Converting object columns to category for better memory management
|
| 43 |
-
# for col in df.select_dtypes(include='object').columns:
|
| 44 |
-
# df[col] = df[col].astype('category')
|
| 45 |
|
| 46 |
-
# Visualization 1:
|
| 47 |
-
st.subheader("
|
| 48 |
|
| 49 |
-
#
|
| 50 |
-
|
|
|
|
| 51 |
|
| 52 |
-
|
| 53 |
-
chart1 = alt.Chart(category_status_counts).mark_bar().encode(
|
| 54 |
x=alt.X('License Type', sort='-y', title='License Type'),
|
| 55 |
-
y=alt.Y('Count', title='Number of Licenses'),
|
| 56 |
-
color=alt.Color('License
|
| 57 |
-
tooltip=['License Type', 'License Status', 'Count']
|
| 58 |
).properties(
|
| 59 |
width=700,
|
| 60 |
height=400,
|
| 61 |
-
title="
|
| 62 |
-
)
|
| 63 |
-
|
| 64 |
st.altair_chart(chart1)
|
| 65 |
|
| 66 |
-
#
|
| 67 |
st.write("""
|
| 68 |
-
**
|
| 69 |
-
|
| 70 |
-
This breakdown can help identify how license types vary not only in volume but also in the proportion of active versus inactive licenses.If I had more time, I might add interactivity to allow users to filter by other variables.
|
| 71 |
""")
|
| 72 |
|
| 73 |
-
# Visualization 2:
|
| 74 |
-
st.subheader("Licenses Over Time")
|
| 75 |
-
|
| 76 |
-
# Ensure the '_id' column exists; otherwise, use another unique identifier if available
|
| 77 |
-
if '_id' in df.columns:
|
| 78 |
-
time_data = df.groupby('Original Issue Year')['_id'].count().reset_index()
|
| 79 |
-
else:
|
| 80 |
-
time_data = df.groupby('Original Issue Year').size().reset_index(name='License Count')
|
| 81 |
|
| 82 |
-
|
|
|
|
| 83 |
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
|
|
|
| 88 |
).properties(
|
| 89 |
-
width=
|
| 90 |
height=400,
|
| 91 |
title="Number of Licenses Issued Over Time"
|
| 92 |
-
)
|
| 93 |
st.altair_chart(chart2)
|
| 94 |
|
| 95 |
-
#
|
| 96 |
st.write("""
|
| 97 |
-
**Licenses Over Time**: This line chart visualizes the trend in the number of licenses issued
|
| 98 |
-
|
| 99 |
-
If I had more time, I would explore monthly or seasonal patterns by adjusting the time grouping.
|
| 100 |
""")
|
|
|
|
| 1 |
+
# import streamlit as st
|
| 2 |
+
# import pandas as pd
|
| 3 |
+
# import altair as alt
|
| 4 |
+
|
| 5 |
+
# # Loading the Dataset
|
| 6 |
+
# url = 'https://github.com/UIUC-iSchool-DataViz/is445_data/raw/main/licenses_fall2022.csv'
|
| 7 |
+
# df = pd.read_csv(url)
|
| 8 |
+
|
| 9 |
+
# # Convert object columns to string for compatibility
|
| 10 |
+
# for col in df.select_dtypes(include=['object']).columns:
|
| 11 |
+
# df[col] = df[col].astype(str)
|
| 12 |
+
|
| 13 |
+
# # Convert integer columns to float
|
| 14 |
+
# for col in df.select_dtypes(include=['int64', 'float64']).columns:
|
| 15 |
+
# df[col] = pd.to_numeric(df[col], errors='coerce')
|
| 16 |
+
|
| 17 |
+
# # Checking for missing values and handling them
|
| 18 |
+
# df['Middle'] = df['Middle'].fillna('N/A') # Update Middle column without inplace
|
| 19 |
+
|
| 20 |
+
# # Dropping columns with more than 90% missing values
|
| 21 |
+
# df.drop(columns=['Title', 'Prefix', 'Suffix', 'BusinessDBA'], inplace=True)
|
| 22 |
+
|
| 23 |
+
# # Dropping rows with missing values for certain key columns
|
| 24 |
+
# df.dropna(subset=['License Number', 'First Name', 'Last Name', 'Effective Date',
|
| 25 |
+
# 'Expiration Date', 'City', 'Zip', 'County', 'Specialty/Qualifier',
|
| 26 |
+
# 'Controlled Substance Schedule', 'Action'], inplace=True)
|
| 27 |
+
|
| 28 |
+
# # Convert datetime columns
|
| 29 |
+
# date_columns = ['Original Issue Date', 'Effective Date', 'Expiration Date', 'LastModifiedDate']
|
| 30 |
+
# for col in date_columns:
|
| 31 |
+
# df[col] = pd.to_datetime(df[col], errors='coerce', format='%m/%d/%Y')
|
| 32 |
+
|
| 33 |
+
# # Add year columns
|
| 34 |
+
# df['Original Issue Year'] = df['Original Issue Date'].dt.year
|
| 35 |
+
# df['Effective Date Year'] = df['Effective Date'].dt.year
|
| 36 |
+
# df['Expiration Date Year'] = df['Expiration Date'].dt.year
|
| 37 |
+
# df['Last Modified Date Year'] = df['LastModifiedDate'].dt.year
|
| 38 |
+
|
| 39 |
+
# # Final check of data types
|
| 40 |
+
# st.write("Column Data Types:", df.dtypes)
|
| 41 |
+
|
| 42 |
+
# # # Converting object columns to category for better memory management
|
| 43 |
+
# # for col in df.select_dtypes(include='object').columns:
|
| 44 |
+
# # df[col] = df[col].astype('category')
|
| 45 |
+
|
| 46 |
+
# # Visualization 1: Licenses by Type with License Status breakdown
|
| 47 |
+
# st.subheader("Licenses by Type and Status")
|
| 48 |
+
|
| 49 |
+
# # Group data by License Type and License Status
|
| 50 |
+
# category_status_counts = df.groupby(['License Type', 'License Status']).size().reset_index(name='Count')
|
| 51 |
+
|
| 52 |
+
# # Create a bar chart with color representing License Status
|
| 53 |
+
# chart1 = alt.Chart(category_status_counts).mark_bar().encode(
|
| 54 |
+
# x=alt.X('License Type', sort='-y', title='License Type'),
|
| 55 |
+
# y=alt.Y('Count', title='Number of Licenses'),
|
| 56 |
+
# color=alt.Color('License Status', legend=alt.Legend(title="License Status")),
|
| 57 |
+
# tooltip=['License Type', 'License Status', 'Count']
|
| 58 |
+
# ).properties(
|
| 59 |
+
# width=700,
|
| 60 |
+
# height=400,
|
| 61 |
+
# title="Number of Licenses by Type and Status"
|
| 62 |
+
# ).interactive()
|
| 63 |
+
|
| 64 |
+
# st.altair_chart(chart1)
|
| 65 |
+
|
| 66 |
+
# # Write-up for Visualization 1
|
| 67 |
+
# st.write("""
|
| 68 |
+
# **Licenses by Type and Status**: This visualization shows the distribution of licenses across different types with further breakdown by license status.
|
| 69 |
+
# The interactive chart allows users to view counts for each license status (e.g., active, inactive) within each license type.
|
| 70 |
+
# This breakdown can help identify how license types vary not only in volume but also in the proportion of active versus inactive licenses.If I had more time, I might add interactivity to allow users to filter by other variables.
|
| 71 |
+
# """)
|
| 72 |
+
|
| 73 |
+
# # Visualization 2: Line chart showing licenses issued over time
|
| 74 |
+
# st.subheader("Licenses Over Time")
|
| 75 |
+
|
| 76 |
+
# # Ensure the '_id' column exists; otherwise, use another unique identifier if available
|
| 77 |
+
# if '_id' in df.columns:
|
| 78 |
+
# time_data = df.groupby('Original Issue Year')['_id'].count().reset_index()
|
| 79 |
+
# else:
|
| 80 |
+
# time_data = df.groupby('Original Issue Year').size().reset_index(name='License Count')
|
| 81 |
+
|
| 82 |
+
# time_data.columns = ['Year', 'License Count']
|
| 83 |
+
|
| 84 |
+
# chart2 = alt.Chart(time_data).mark_line().encode(
|
| 85 |
+
# x=alt.X('Year', title='Year'),
|
| 86 |
+
# y=alt.Y('License Count', title='Number of Licenses'),
|
| 87 |
+
# color=alt.value('blue')
|
| 88 |
+
# ).properties(
|
| 89 |
+
# width=600,
|
| 90 |
+
# height=400,
|
| 91 |
+
# title="Number of Licenses Issued Over Time"
|
| 92 |
+
# )
|
| 93 |
+
# st.altair_chart(chart2)
|
| 94 |
+
|
| 95 |
+
# # Write-up for Visualization 2
|
| 96 |
+
# st.write("""
|
| 97 |
+
# **Licenses Over Time**: This line chart visualizes the trend in the number of licenses issued over time.
|
| 98 |
+
# I chose a line chart to emphasize changes over years. Blue was used to create a consistent look. Axis labels and title are provided for clarity.
|
| 99 |
+
# If I had more time, I would explore monthly or seasonal patterns by adjusting the time grouping.
|
| 100 |
+
# """)
|
| 101 |
+
|
| 102 |
+
|
| 103 |
import streamlit as st
|
| 104 |
import pandas as pd
|
| 105 |
import altair as alt
|
|
|
|
| 108 |
url = 'https://github.com/UIUC-iSchool-DataViz/is445_data/raw/main/licenses_fall2022.csv'
|
| 109 |
df = pd.read_csv(url)
|
| 110 |
|
| 111 |
+
# Dropping problematic columns
|
| 112 |
+
df.drop(columns=['_id', 'Case Number', 'Delegated Controlled Substance Schedule'], inplace=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 113 |
|
| 114 |
+
# Handling missing data
|
| 115 |
+
df['Middle'].fillna('N/A', inplace=True)
|
| 116 |
+
df.dropna(subset=['License Number', 'First Name', 'Last Name', 'Effective Date', 'Expiration Date', 'City', 'Zip', 'County', 'Specialty/Qualifier', 'Controlled Substance Schedule', 'Action'], inplace=True)
|
| 117 |
|
| 118 |
+
# Converting date columns to datetime format
|
| 119 |
+
df['Original Issue Date'] = pd.to_datetime(df['Original Issue Date'], errors='coerce', format='%m/%d/%Y')
|
| 120 |
+
df['Effective Date'] = pd.to_datetime(df['Effective Date'], errors='coerce', format='%m/%d/%Y')
|
| 121 |
+
df['Expiration Date'] = pd.to_datetime(df['Expiration Date'], errors='coerce', format='%m/%d/%Y')
|
| 122 |
+
df['LastModifiedDate'] = pd.to_datetime(df['LastModifiedDate'], errors='coerce', format='%m/%d/%Y')
|
| 123 |
|
| 124 |
+
# Adding Year columns
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 125 |
df['Original Issue Year'] = df['Original Issue Date'].dt.year
|
| 126 |
df['Effective Date Year'] = df['Effective Date'].dt.year
|
| 127 |
df['Expiration Date Year'] = df['Expiration Date'].dt.year
|
| 128 |
df['Last Modified Date Year'] = df['LastModifiedDate'].dt.year
|
| 129 |
|
| 130 |
+
# Converting object columns to category
|
| 131 |
+
for col in df.select_dtypes(include='object').columns:
|
| 132 |
+
df[col] = df[col].astype('category')
|
|
|
|
|
|
|
|
|
|
| 133 |
|
| 134 |
+
# Visualization 1: Distribution of License Types (Enhanced)
|
| 135 |
+
st.subheader("Distribution of License Types (Log Scale)")
|
| 136 |
|
| 137 |
+
# Aggregating and plotting data
|
| 138 |
+
license_type_counts = df['License Type'].value_counts().reset_index()
|
| 139 |
+
license_type_counts.columns = ['License Type', 'Count']
|
| 140 |
|
| 141 |
+
chart1 = alt.Chart(license_type_counts).mark_bar().encode(
|
|
|
|
| 142 |
x=alt.X('License Type', sort='-y', title='License Type'),
|
| 143 |
+
y=alt.Y('Count', title='Number of Licenses (Log Scale)', scale=alt.Scale(type='log')),
|
| 144 |
+
color=alt.Color('License Type', legend=None)
|
|
|
|
| 145 |
).properties(
|
| 146 |
width=700,
|
| 147 |
height=400,
|
| 148 |
+
title="Distribution of License Types (Log Scale)"
|
| 149 |
+
)
|
|
|
|
| 150 |
st.altair_chart(chart1)
|
| 151 |
|
| 152 |
+
# Writeup for Visualization 1
|
| 153 |
st.write("""
|
| 154 |
+
**Distribution of License Types**: This bar chart (using a log scale for clarity) displays the distribution of different license types.
|
| 155 |
+
Due to the imbalance in the dataset, a logarithmic scale is applied to better represent both common and rare license types.
|
|
|
|
| 156 |
""")
|
| 157 |
|
| 158 |
+
# Visualization 2: Trend of Licenses Issued Over Time (Enhanced)
|
| 159 |
+
st.subheader("Trend of Licenses Issued Over Time (Interactive)")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 160 |
|
| 161 |
+
# Aggregating yearly data
|
| 162 |
+
time_data = df.groupby('Original Issue Year').size().reset_index(name='License Count')
|
| 163 |
|
| 164 |
+
# Plotting the trend with additional interactivity
|
| 165 |
+
chart2 = alt.Chart(time_data).mark_line(point=True).encode(
|
| 166 |
+
x=alt.X('Original Issue Year:O', title='Year'),
|
| 167 |
+
y=alt.Y('License Count', title='Number of Licenses Issued'),
|
| 168 |
+
tooltip=['Original Issue Year', 'License Count']
|
| 169 |
).properties(
|
| 170 |
+
width=700,
|
| 171 |
height=400,
|
| 172 |
title="Number of Licenses Issued Over Time"
|
| 173 |
+
).interactive()
|
| 174 |
st.altair_chart(chart2)
|
| 175 |
|
| 176 |
+
# Writeup for Visualization 2
|
| 177 |
st.write("""
|
| 178 |
+
**Trend of Licenses Issued Over Time**: This line chart visualizes the trend in the number of licenses issued each year.
|
| 179 |
+
Interactive points allow viewers to see the specific count for each year. This chart provides insight into trends in licensing activity over time.
|
|
|
|
| 180 |
""")
|