Skip to content

Latest commit

 

History

History
1741 lines (1502 loc) · 42.3 KB

FoodAccessAnalysis.md

File metadata and controls

1741 lines (1502 loc) · 42.3 KB

Food Access Analysis

Setup and Data Review

# Import libraries 
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

# Set Vis style
sns.set_style("whitegrid")
plt.rcParams["figure.figsize"] = (10, 6)

# Load the cleaned data
path = r'C:\Users\mahfo\Documents\portfolio\Data\FD3.csv'
FD = pd.read_csv(path)
# Display basic info about the dataset
print("Dataset Info:")
print(FD.info())

# Preview the first few rows
print("\nFirst 5 Rows of Dataset:")
print(FD.head())
Dataset Info:
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 72531 entries, 0 to 72530
Data columns (total 27 columns):
 #   Column                      Non-Null Count  Dtype  
---  ------                      --------------  -----  
 0   Tract_ID                    72531 non-null  int64  
 1   fipscode                    72531 non-null  int64  
 2   State                       72531 non-null  object 
 3   County                      72531 non-null  object 
 4   Urban_Status                72531 non-null  object 
 5   LowFoodAccess               72531 non-null  object 
 6   Total_Population            72531 non-null  int64  
 7   LowAccess_Pop               72531 non-null  float64
 8   LowIncome_Pop               72531 non-null  float64
 9   LowIncome&_LowAccess_Pop    72531 non-null  float64
 10  NoVehicle_HHs               72531 non-null  float64
 11  SNAP_HHs                    72531 non-null  float64
 12  Poverty_Rate                72531 non-null  float64
 13  Median_Income               72531 non-null  float64
 14  LowAccess_Ratio             72531 non-null  float64
 15  LowIncome&_LowAccess_Ratio  72531 non-null  float64
 16  Pct_LowIncome_              72531 non-null  float64
 17  NoVehicle_HHs_Ratio         72531 non-null  float64
 18  SNAP_Participatio_Rate      72531 non-null  float64
 19  premature_death             72531 non-null  float64
 20  poor_health                 72531 non-null  float64
 21  low_birthweight             72531 non-null  float64
 22  obesity                     72531 non-null  float64
 23  diabetes                    72531 non-null  float64
 24  unemployment                72531 non-null  float64
 25  child_poverty               72531 non-null  float64
 26  income_inequality           72531 non-null  float64
dtypes: float64(20), int64(3), object(4)
memory usage: 14.9+ MB
None

First 5 Rows of Dataset:
     Tract_ID  fipscode    State          County Urban_Status LowFoodAccess  \
0  1001020100      1001  Alabama  Autauga County        Urban            NO   
1  1001020200      1001  Alabama  Autauga County        Urban           YES   
2  1001020300      1001  Alabama  Autauga County        Urban            NO   
3  1001020400      1001  Alabama  Autauga County        Urban            NO   
4  1001020500      1001  Alabama  Autauga County        Urban            NO   

   Total_Population  LowAccess_Pop  LowIncome_Pop  LowIncome&_LowAccess_Pop  \
0              1912         1896.0          455.0                     461.0   
1              2170         1261.0          802.0                     604.0   
2              3373         1552.0         1306.0                     478.0   
3              4386         1363.0          922.0                     343.0   
4             10766         2643.0         2242.0                     586.0   

   ...  NoVehicle_HHs_Ratio  SNAP_Participatio_Rate  premature_death  \
0  ...             0.003138                0.053347           9778.1   
1  ...             0.041014                0.071889           9778.1   
2  ...             0.029351                0.050993           9778.1   
3  ...             0.004788                0.022344           9778.1   
4  ...             0.021364                0.031488           9778.1   

   poor_health  low_birthweight  obesity  diabetes  unemployment  \
0       0.2586           0.0883      0.3    0.8149         0.046   
1       0.2586           0.0883      0.3    0.8149         0.046   
2       0.2586           0.0883      0.3    0.8149         0.046   
3       0.2586           0.0883      0.3    0.8149         0.046   
4       0.2586           0.0883      0.3    0.8149         0.046   

   child_poverty  income_inequality  
0          0.138               40.7  
1          0.138               40.7  
2          0.138               40.7  
3          0.138               40.7  
4          0.138               40.7  

[5 rows x 27 columns]

# Preview the first few rows
print("\nFirst 5 Rows of Dataset:")
print(FD.head())
First 5 Rows of Dataset:
     Tract_ID  fipscode    State          County Urban_Status LowFoodAccess  \
0  1001020100      1001  Alabama  Autauga County        Urban            NO   
1  1001020200      1001  Alabama  Autauga County        Urban           YES   
2  1001020300      1001  Alabama  Autauga County        Urban            NO   
3  1001020400      1001  Alabama  Autauga County        Urban            NO   
4  1001020500      1001  Alabama  Autauga County        Urban            NO   

   Total_Population  LowAccess_Pop  LowIncome_Pop  LowIncome&_LowAccess_Pop  \
0              1912         1896.0          455.0                     461.0   
1              2170         1261.0          802.0                     604.0   
2              3373         1552.0         1306.0                     478.0   
3              4386         1363.0          922.0                     343.0   
4             10766         2643.0         2242.0                     586.0   

   ...  NoVehicle_HHs_Ratio  SNAP_Participatio_Rate  premature_death  \
0  ...             0.003138                0.053347           9778.1   
1  ...             0.041014                0.071889           9778.1   
2  ...             0.029351                0.050993           9778.1   
3  ...             0.004788                0.022344           9778.1   
4  ...             0.021364                0.031488           9778.1   

   poor_health  low_birthweight  obesity  diabetes  unemployment  \
0       0.2586           0.0883      0.3    0.8149         0.046   
1       0.2586           0.0883      0.3    0.8149         0.046   
2       0.2586           0.0883      0.3    0.8149         0.046   
3       0.2586           0.0883      0.3    0.8149         0.046   
4       0.2586           0.0883      0.3    0.8149         0.046   

   child_poverty  income_inequality  
0          0.138               40.7  
1          0.138               40.7  
2          0.138               40.7  
3          0.138               40.7  
4          0.138               40.7  

[5 rows x 27 columns]

print(FD.columns)
Index(['Tract_ID', 'fipscode', 'State', 'County', 'Urban_Status',
       'LowFoodAccess', 'Total_Population', 'LowAccess_Pop', 'LowIncome_Pop',
       'LowIncome&_LowAccess_Pop', 'NoVehicle_HHs', 'SNAP_HHs', 'Poverty_Rate',
       'Median_Income', 'LowAccess_Ratio', 'LowIncome&_LowAccess_Ratio',
       'Pct_LowIncome_', 'NoVehicle_HHs_Ratio', 'SNAP_Participatio_Rate',
       'premature_death', 'poor_health', 'low_birthweight', 'obesity',
       'diabetes', 'unemployment', 'child_poverty', 'income_inequality'],
      dtype='object')

#Drop these columns 'LowAccess_Ratio', 'LowIncome&_LowAccess_Ratio', 'Pct_LowIncome_', 'NoVehicle_HHs_Ratio', 'SNAP_Participatio_Rate',
FD = FD[['Tract_ID', 'fipscode', 'State', 'County', 
         'Urban_Status','LowFoodAccess',
         'Total_Population', 'LowAccess_Pop', 'LowIncome_Pop',
       'LowIncome&_LowAccess_Pop',
       'NoVehicle_HHs', 'SNAP_HHs',  'Poverty_Rate', 'Median_Income',
       'premature_death', 'poor_health', 'low_birthweight', 'obesity',
       'diabetes', 'unemployment', 'child_poverty', 'income_inequality']]

Overview of the Data to give us an idea

Such as food access numbers , uraban vs rural. This is not good for analyis or answering question but rtahter an over view

# Urban vs. Rural Count

Urban_count = FD.groupby('Urban_Status')[['LowFoodAccess']].count()
print("Urban_Status Count:")
print(Urban_count)

#Counties by Urban vs. Rural
sns.countplot(data=FD, x="Urban_Status", hue="Urban_Status")
plt.title("Distribution of Counties by Urban vs. Rural")
plt.xlabel(None)
plt.ylabel("Number of Counties")
plt.show()

Urban_Status Count:
              LowFoodAccess
Urban_Status               
Rural                 17362
Urban                 55169

png

FoodAccess_count = FD.groupby('LowFoodAccess')[['Urban_Status']].count()
print("Food Access by Urban Status:")
print(Urban_count)
# Food Access Classification
sns.countplot(data=FD, x="LowFoodAccess", hue="LowFoodAccess")
plt.title("Counties with Low Food Access")
plt.xlabel("Low Food Access")
plt.ylabel("Number of Counties")
plt.show()
Food Access by Urban Status:
              LowFoodAccess
Urban_Status               
Rural                 17362
Urban                 55169

png

#2x2 table (contingency table / cross-tabulation)

Urban_contingency_table = pd.crosstab(FD["LowFoodAccess"], FD["Urban_Status"])

print("Contingency Table (LowFoodAccess vs Urban_Status):")
print(Urban_contingency_table)
Contingency Table (LowFoodAccess vs Urban_Status):
Urban_Status   Rural  Urban
LowFoodAccess              
NO             16010  47228
YES             1352   7941

plt.figure(figsize=(8,5))
sns.countplot(data=FD, x='Urban_Status', hue='LowFoodAccess', 
             palette={'YES':'salmon', 'NO':'lightgreen'})
plt.title('Food Deserts in Urban vs Rural Areas')
plt.xlabel(None)
plt.ylabel('Frequency')
plt.legend(title='Food Desert')
plt.show()

png

Defining Key Questions to Answer

Food Access and Population Impact

Which states have the largest populations affected by low food access, and what proportion of their population is impacted? (Building on previous work) How does low food access vary between urban and rural areas in terms of affected population and proportion?

Socioeconomic Factors

Is there a relationship between low food access and poverty or income inequality? How does vehicle ownership (or lack thereof) correlate with low food access?

Health Outcomes

Are areas with low food access associated with worse health outcomes, such as higher obesity or diabetes rates? Is there a link between low food access and premature death rates?

Geographic and Demographic Patterns

Which counties within a specific state (e.g., top-affected state) have the most severe food access issues? Does SNAP (food assistance) participation correlate with low food access or low-income populations? These questions will guide our analysis, helping us uncover actionable insights while demonstrating a range of data analysis techniques.

Population-Weighted Analysis of Food Access

Question 1: States with Largest Affected Populations and Proportions

# Filter low food access(LFA) tracts
LFA_tract = FD[FD["LowFoodAccess"] == "YES"]
# Total population with low food access(LFA) by state
FD["Total_Population"]
LFA_by_state = LFA_tract.groupby("State")["Total_Population"].sum()
#Total pop by state
total_pop_by_state = FD.groupby("State")["Total_Population"].sum()
# Proportion of pop with low food access(LFA)
LFA_prop = (LFA_by_state/total_pop_by_state*100).sort_values(ascending=False).round(decimals=1)
LFA_tract
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
Tract_ID fipscode State County Urban_Status LowFoodAccess Total_Population LowAccess_Pop LowIncome_Pop LowIncome&_LowAccess_Pop ... NoVehicle_HHs_Ratio SNAP_Participatio_Rate premature_death poor_health low_birthweight obesity diabetes unemployment child_poverty income_inequality
1 1001020200 1001 Alabama Autauga County Urban YES 2170 1261.0 802.0 604.0 ... 0.041014 0.071889 9778.1 0.2586 0.088300 0.300 0.8149 0.046 0.138 40.7
5 1001020600 1001 Alabama Autauga County Urban YES 3668 3438.0 1659.0 1585.0 ... 0.019357 0.061069 9778.1 0.2586 0.088300 0.300 0.8149 0.046 0.138 40.7
6 1001020700 1001 Alabama Autauga County Urban YES 2891 1231.0 2175.0 742.0 ... 0.011761 0.134901 9778.1 0.2586 0.088300 0.300 0.8149 0.046 0.138 40.7
10 1001021000 1001 Alabama Autauga County Rural YES 2894 2338.0 977.0 902.0 ... 0.003110 0.050449 9778.1 0.2586 0.088300 0.300 0.8149 0.046 0.138 40.7
11 1001021100 1001 Alabama Autauga County Rural YES 3320 2640.0 1463.0 1354.0 ... 0.081024 0.069277 9778.1 0.2586 0.088300 0.300 0.8149 0.046 0.138 40.7
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
72455 56021000700 56021 Wyoming Laramie County Urban YES 4338 4162.0 1772.0 2004.0 ... 0.101890 0.063624 7837.0 0.1187 0.094000 0.242 0.8055 0.039 0.118 39.9
72465 56021001502 56021 Wyoming Laramie County Urban YES 4899 1627.0 1857.0 601.0 ... 0.032456 0.050623 7837.0 0.1187 0.094000 0.242 0.8055 0.039 0.118 39.9
72473 56025000200 56025 Wyoming Natrona County Urban YES 4385 4137.0 1946.0 2078.0 ... 0.074116 0.080730 8175.6 0.1301 0.072000 0.250 0.7888 0.029 0.137 42.3
72491 56027957200 56027 Wyoming Niobrara County Rural YES 2484 545.0 941.0 230.0 ... 0.012882 0.022142 10210.6 0.0882 0.080252 0.268 0.7826 0.038 0.164 43.3
72497 56031959100 56031 Wyoming Platte County Rural YES 2092 630.0 758.0 180.0 ... 0.014818 0.028203 7985.4 0.1071 0.092900 0.224 0.7692 0.041 0.190 41.3

9293 rows × 27 columns

LFA_by_state
total_pop_by_state
LFA_prop
State
Mississippi             29.8
New Mexico              26.6
Arkansas                24.0
Louisiana               22.3
Georgia                 22.3
Texas                   19.6
Alabama                 19.0
South Carolina          18.9
Missouri                18.1
Tennessee               17.9
Indiana                 17.5
Kansas                  17.2
North Carolina          16.7
Arizona                 16.4
Oklahoma                16.1
South Dakota            14.8
Delaware                14.5
Virginia                14.3
Minnesota               13.9
Alaska                  13.6
Florida                 13.5
West Virginia           13.5
Colorado                13.2
Kentucky                13.2
New Hampshire           13.1
Ohio                    13.0
Oregon                  12.9
Idaho                   12.5
Montana                 12.3
Washington              12.3
Michigan                11.4
Nebraska                11.0
Hawaii                  10.3
Iowa                    10.3
Illinois                 9.7
Maryland                 9.6
Wisconsin                8.9
Connecticut              8.9
Wyoming                  8.8
Utah                     8.5
Massachusetts            8.3
North Dakota             7.9
Maine                    7.6
Vermont                  7.4
California               7.2
Nevada                   7.0
Pennsylvania             6.3
District of Columbia     5.9
Rhode Island             5.4
New Jersey               5.2
New York                 3.9
Name: Total_Population, dtype: float64

# Combine into a single dataframe
state_summary = pd.DataFrame({
    "LFA_Population" : LFA_by_state,
    "Total_Population" : total_pop_by_state,
    "LFA_Percentage (%)": LFA_prop
}).sort_values(by="LFA_Population", ascending=False)
state_summary
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
LFA_Population Total_Population LFA_Percentage (%)
State
Texas 4926344 25145561 19.6
California 2669879 37253956 7.2
Florida 2546335 18801310 13.5
Georgia 2158463 9687653 22.3
North Carolina 1593822 9535483 16.7
Ohio 1504341 11536504 13.0
Illinois 1242939 12830632 9.7
Virginia 1147233 8001024 14.3
Indiana 1135595 6483802 17.5
Tennessee 1134333 6346105 17.9
Michigan 1131629 9883640 11.4
Missouri 1084564 5988927 18.1
Arizona 1049466 6392017 16.4
Louisiana 1010836 4533372 22.3
Alabama 908205 4779736 19.0
Mississippi 884934 2967297 29.8
South Carolina 872233 4625364 18.9
Washington 825837 6724540 12.3
Pennsylvania 800303 12702379 6.3
New York 757797 19378102 3.9
Minnesota 737593 5303925 13.9
Arkansas 698688 2915918 24.0
Colorado 663246 5029196 13.2
Oklahoma 602480 3751351 16.1
Kentucky 571751 4339367 13.2
Maryland 552017 5773552 9.6
New Mexico 547468 2059179 26.6
Massachusetts 541476 6547629 8.3
Wisconsin 505977 5686986 8.9
Oregon 494475 3831074 12.9
Kansas 491894 2853118 17.2
New Jersey 459081 8791894 5.2
Connecticut 317446 3574097 8.9
Iowa 313703 3046355 10.3
West Virginia 250113 1852994 13.5
Utah 234217 2763885 8.5
Nebraska 200184 1826341 11.0
Idaho 195299 1567582 12.5
Nevada 188652 2700551 7.0
New Hampshire 172797 1316470 13.1
Hawaii 140491 1360301 10.3
Delaware 129852 897934 14.5
Montana 122057 989415 12.3
South Dakota 120657 814180 14.8
Maine 100822 1328361 7.6
Alaska 96841 710231 13.6
Rhode Island 56721 1052567 5.4
North Dakota 52813 672591 7.9
Wyoming 49592 563626 8.8
Vermont 46079 625741 7.4
District of Columbia 35404 601723 5.9 
print(f"Top 10 States by Low food Access Population and Proportion: \n{state_summary.head(10)}")
Top 10 States by Low food Access Population and Proportion: 
                LFA_Population  Total_Population  LFA_Percentage (%)
State                                                               
Texas                  4926344          25145561                19.6
California             2669879          37253956                 7.2
Florida                2546335          18801310                13.5
Georgia                2158463           9687653                22.3
North Carolina         1593822           9535483                16.7
Ohio                   1504341          11536504                13.0
Illinois               1242939          12830632                 9.7
Virginia               1147233           8001024                14.3
Indiana                1135595           6483802                17.5
Tennessee              1134333           6346105                17.9

# Bar plot for population
top_10_states = state_summary.head(10)
plt.figure(figsize=(12, 6))
sns.barplot(data=top_10_states, x = "State", y = "LFA_Population")
plt.xlabel("Sate")
plt.ylabel("Number of People with Low Food Access (in Millions)")
plt.title("Top 10 States by Population with Low Food Access")
Text(0.5, 1.0, 'Top 10 States by Population with Low Food Access')

png

Create a twin Axes sharing the xaxis.

Create a new Axes with an invisible x-axis and an independent y-axis positioned opposite to the original one (i.e. at right). The x-axis autoscale setting will be inherited from the original Axes. To ensure that the tick marks of both y-axes align, see ~matplotlib.ticker.LinearLocator.

state_summary
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
LFA_Population Total_Population LFA_Percentage (%)
State
Texas 4926344 25145561 19.6
California 2669879 37253956 7.2
Florida 2546335 18801310 13.5
Georgia 2158463 9687653 22.3
North Carolina 1593822 9535483 16.7
Ohio 1504341 11536504 13.0
Illinois 1242939 12830632 9.7
Virginia 1147233 8001024 14.3
Indiana 1135595 6483802 17.5
Tennessee 1134333 6346105 17.9
Michigan 1131629 9883640 11.4
Missouri 1084564 5988927 18.1
Arizona 1049466 6392017 16.4
Louisiana 1010836 4533372 22.3
Alabama 908205 4779736 19.0
Mississippi 884934 2967297 29.8
South Carolina 872233 4625364 18.9
Washington 825837 6724540 12.3
Pennsylvania 800303 12702379 6.3
New York 757797 19378102 3.9
Minnesota 737593 5303925 13.9
Arkansas 698688 2915918 24.0
Colorado 663246 5029196 13.2
Oklahoma 602480 3751351 16.1
Kentucky 571751 4339367 13.2
Maryland 552017 5773552 9.6
New Mexico 547468 2059179 26.6
Massachusetts 541476 6547629 8.3
Wisconsin 505977 5686986 8.9
Oregon 494475 3831074 12.9
Kansas 491894 2853118 17.2
New Jersey 459081 8791894 5.2
Connecticut 317446 3574097 8.9
Iowa 313703 3046355 10.3
West Virginia 250113 1852994 13.5
Utah 234217 2763885 8.5
Nebraska 200184 1826341 11.0
Idaho 195299 1567582 12.5
Nevada 188652 2700551 7.0
New Hampshire 172797 1316470 13.1
Hawaii 140491 1360301 10.3
Delaware 129852 897934 14.5
Montana 122057 989415 12.3
South Dakota 120657 814180 14.8
Maine 100822 1328361 7.6
Alaska 96841 710231 13.6
Rhode Island 56721 1052567 5.4
North Dakota 52813 672591 7.9
Wyoming 49592 563626 8.8
Vermont 46079 625741 7.4
District of Columbia 35404 601723 5.9 
# Dual-axis visualization for top 10 states by population
# Create figure and primary axes
#fig, ax1 = plt.subplots() 
fig, ax1 = plt.subplots()
sns.barplot(data=top_10_states, x = "State", y = "LFA_Population", ax = ax1, color="blue")
ax1.set_ylabel("Number of People with Low Food Access(Millions)", color = "blue")
ax2 = ax1.twinx()
sns.lineplot(data=top_10_states, x = "State", y=top_10_states["LFA_Percentage (%)"], ax= ax2, color = "red",marker="o")
ax2.set_ylabel("Percentage of People with Low Food Access", color ="red")
plt.title("Top 10 State: Low food Access")
fig.tight_layout()
plt.show()

png

Question 2: Urban vs. Rural Impact

#  Filter low food access(LFA) tracts
# LFA_tract = FD[FD["LowFoodAccess"] == "YES"]
# Population with low food access by urban status
# Total population with low food access(LFA) by urban status
LFA_Pop_by_UrbanStatus = LFA_tract.groupby("Urban_Status")["Total_Population"].sum()
#Total pop by Urban_Status
Total_Pop_by_UrbanStatus = FD.groupby("Urban_Status")["Total_Population"].sum()
# Proportion of pop with low food access(LFA)  based on Urban_Status
LFA_Prop_Urban = (LFA_Pop_by_UrbanStatus/Total_Pop_by_UrbanStatus*100)

# Combine into a single dataframe
urban_summary = pd.DataFrame({
    "Low_Food_Access_Pop" : LFA_Pop_by_UrbanStatus,
    'Proportion (%)': LFA_Prop_Urban
})
# Display combined results
print("Urban vs. Rural Low Food Access:")
print(urban_summary)
# Stacked bar plot with proportion annotation
fig, ax1 = plt.subplots()
sns.barplot(data=urban_summary, x = "Urban_Status", y = "Low_Food_Access_Pop", hue="Urban_Status", ax =ax1)
ax1.set_ylabel("Number of People with Low Food Access (10 X Millions)")
ax1.set(xlabel=None)
ax2=ax1.twinx()
sns.lineplot(data=urban_summary, x = "Urban_Status", y = urban_summary["Proportion (%)"], ax =ax2,color = "darkgreen", marker="o")
ax2.set_ylabel("Percentage of People with Low Food Access", color = "darkgreen")
Urban vs. Rural Low Food Access:
              Low_Food_Access_Pop  Proportion (%)
Urban_Status                                     
Rural                     4673301        6.584775
Urban                    34401673       14.468209





Text(0, 0.5, 'Percentage of People with Low Food Access')

png

LFA_Pop_by_UrbanStatus
Total_Pop_by_UrbanStatus
Urban_Status
Rural     70971304
Urban    237774234
Name: Total_Population, dtype: int64

Socioeconomic Analysis & Health Outcomes

Question 3: Food Access: Probable Causes and Consequences (Correlation Analysis)

#Add ratios columns 
FD["LFA_Rate"] = FD["LowAccess_Pop"] / FD["Total_Population"] * 100
FD["SNAP_Rate"] = FD["SNAP_HHs"] / FD["Total_Population"] * 100
FD["NoVehicle_Rate"] = FD["NoVehicle_HHs"] / FD["Total_Population"] * 100


# Select food access (LFA_Rate), its probable causes, and consequences columns
FD.columns
numeric_cols = FD[['Poverty_Rate',
       'Median_Income', 'premature_death', 'poor_health', 'low_birthweight',
       'obesity', 'diabetes', 'unemployment', 'child_poverty',
       'income_inequality', 'LFA_Rate', 'SNAP_Rate', 'NoVehicle_Rate']]

# Compute correlation matrix
corr = numeric_cols.corr()

# Plot heatmap
plt.figure(figsize=(12, 8))
sns.heatmap(corr, annot=True, fmt=".2f", cmap="coolwarm", linewidths=0.5)
plt.title("Correlation Heatmap of Food Access Variables")
plt.show()

png

Goal: Understand whether low food access (LFA_Rate) is linked to poverty, transportation, and health.

Method: I used Pearson correlation to test relationships between food access and potential causes (e.g., Poverty_Rate, NoVehicle_Rate) and effects (e.g., obesity).

Findings:

LFA_Rate had very weak or no correlation with poverty, SNAP participation, vehicle access, or obesity.

The strongest relationship found was between:

SNAP_Rate and Poverty_Rate: +0.69

child_poverty and poor_health: +0.71

These reflect clear socioeconomic links, even if food access alone doesn't show up as a major player. Even though we often assume food deserts lead directly to poor health or are caused by poverty and lack of transportation, this dataset shows that simple linear relationships are limited. Food access is a complex issue influenced by multiple overlapping factors, and correlation alone doesn’t capture the full story.

Health Outcomes

Linking Childhood Poverty to Adult Health: Urban and Rural Perspectives

The scatter plot explores the relationship between two variables: Proportion of Adults in Poor Health (x-axis) : This represents the percentage of adults in a given area or population who are in poor health. Proportion of Children in Poverty (y-axis) : This represents the percentage of children in the same area or population who are living in poverty.

# Poor Health vs Child Poverty
sns.scatterplot(data=FD, x="poor_health", y="child_poverty", hue="Urban_Status")
plt.title("Poor Health vs Child Poverty")
plt.xlabel("Proportion of Adults Health Status")
plt.ylabel("Proportion of Children in Poverty ")
plt.show()

png

Question 4: Low Food Access VS. Obesity and Diabetes

# Average health metrics
health_comp = FD.groupby('LowFoodAccess')[['obesity', 'diabetes']].mean()
print("Health Metrics by Food Access:")
print(health_comp)

# Boxplot for obesity
sns.boxplot(data=FD, x='LowFoodAccess', y='obesity', hue="LowFoodAccess")
plt.title('Obesity by Low Food Access')
plt.show()
Health Metrics by Food Access:
                obesity  diabetes
LowFoodAccess                    
NO             0.257326  0.796038
YES            0.276815  0.791431

png

# Boxplot for diabetes
sns.boxplot(data=FD, x='LowFoodAccess', y='diabetes', hue="LowFoodAccess")
plt.xlabel("Low Food Access Status")
plt.ylabel("Diabety Rate")
plt.title('Diabetes by Low Food Access')
plt.show()

png

Geographic Drill-Down

Question 5: County-Level View in Most Affected State

# Get top state and filter data
top_state = state_summary.index[0]
state_data = low_access_tracts[low_access_tracts['State'] == top_state]
county_data = state_data.groupby('County')['Total_Population'].sum().sort_values(ascending=False).head(10)

print(f"Top 10 Counties in {top_state}:")
print(county_data)
Top 10 Counties in Texas:
County
Harris County      637644
Hidalgo County     461260
Dallas County      441656
Tarrant County     315695
Bexar County       315428
El Paso County     205339
Cameron County     184305
Travis County      135025
Bell County         82490
McLennan County     68357
Name: Total_Population, dtype: int64

# Generate gradient colors
n_bars = len(county_data)
colors = plt.cm.viridis(np.linspace(0, 1, n_bars)).tolist()  # Convert to list

# Create the plot with correct parameters
ax = sns.barplot(
    x=county_data.values,
    y=county_data.index,
    hue=county_data.index,  # Required now
    palette=colors
)

plt.title(f"Top Counties in {top_state}")
plt.xlabel("Population")
plt.tight_layout()
plt.show()

png

poor_health VS child_poverty

poor_health: Proportion of adults who report being in fair or poor general health, expressed as a value between 0 and 1 (i.e., 0 = no one reports poor health, 1 = everyone reports poor health). child_poverty: Proportion of children under age 18 living in households below the federal poverty threshold, expressed as a value between 0 and 1.

# Poor Health vs Child Poverty
sns.scatterplot(data=FD, x="poor_health", y="child_poverty", hue="Urban_Status")
plt.title("Poor Health vs Child Poverty")
plt.xlabel("Proportion of Adults Health Status")
plt.ylabel("Proportion of Children in Poverty ")
plt.show()

png

There is a positive correlation between poor health and child poverty, with urban areas showing more variability and rural areas having distinct patterns.