Skip to content

Repository files navigation

project-spring24: MCA

Team members

  1. TaoYuan(lead) github username:hueyyt
  2. Qian Yin github username:PennyY1103
  3. Sihan Bai github username:hanasu061

Data source

  • According to the Purdue's 2021 Digital Divide Index, we need two parts: INFA and SE.

  • INFA data comes from the FCC broadband data collection.

  • SE data comes from the American Community Survey 5-Year Data (2009-2022).

    • We used detailed tables, subject tables, data profiles and comparsion profiles.

    • Detailed Tables contain the most detailed cross-tabulations, many of which are published down to block groups. The data are population counts.

    • Subject Tables provide an overview of the estimates available in a particular topic. The data are presented as population counts and percentages.

    • Maine's number in the table: 23

    • Persons who are 60 years of age or older (Subject tables)

    • Incarcerated individuals

      • Used 2022 State Data in Digital Equity Act of 2021 (census.gov).
      • It has total population of Maine state and percentage of incarcerated population.
      • estimated incarcerated population in Maine = state_tot_pop * pct_incarc_pop
      • proportion = num of people in each tract / total population in Maine
      • estimated incarcerated population in each tract = estimated incarcerated population in Maine * proportion
    • Veterans (Detailed table)

    • Persons with disabilities (Subject table)

    • Members of a racial or ethnic minority group

      • (def) The population who identify as a race other than White alone or as Hispanic or Latino of any race
      • (data source) American Community Survey B01001(Sex by Age) 2022 5-Year Estimates Detailed Tables
    • Rural residents

      • (def) The rural population, as defined by the Digital Equity Act of 2021, is the population living outside of cities and towns with more than 20,000 residents or living outside larger cities and towns with more than 50,000 residents and their surrounding urban areas.
      • (data source) Digital Equity Act of 2021 2022 State Data, 2019 County and Census Tract Data, and American Community Survey P2(Urban and Rural) 2020 Congressional District Summary File
    • Individuals with a language barrier, including those who are English learners or have low literacy levels

      • (def) English learners or have low literacy levels. English learners are defined as individuals who speak a language other than English at home and speak English less than “very well.”
      • (data source) Digital Equity Act of 2021 2022 State Data
    • Individuals living in households with incomes not exceeding 150 percent of the poverty level

      • (def) Households with income no more than 150 percent of the federal poverty threshold.
      • American Community Survey S1701(Poverty Status in the Past 12 Months) 2022 5-Year Estimates Subject Tables

Data processing

INFA (infrastructure/adoption)

In order to print the INFA results, please use the following commands:

make data

This command will download data about INFA.

make speed

This is for calculate average downlaod and upload speed. It will output a file named speed.txt.

make population

For calculate percentage of populaiton can't access to network or compute device. It will output a file named population_pct_analysis.txt.

make INFA

To calculate INFA. It will output a file named infa_scaled.csv. This is the first head of output.

              id  INFA_scaled
0    23001010100   100.000000
1    23001010100   100.000000
2    23001010200     7.227720
3    23001010200     7.227720
4    23001010300    46.613047
make maps

To combine INFA and geo data and make maps by geo id. It will output a file named INFA_scaled_tract.png.

SE (socioeconomic)

First of all, the openpyxl package needs to be installed so that excel files can be read. The following command is used to install openpyxl. conda install openpyxl

To calculate the estimated rural residents and save the result into a .csv file, use the following command:

make rural_est
  • The output of this command is list below. It includes the estimation of rural residents. Although the result of the first approach is the closest one, the estimated distribution can be highly inaccurate. Therefore, the second approach is selected.
Total rural population in maine (actual):
 1051000
Total rural population in maine (estimation v1):
 1037051
Total rural population in maine (estimation v2):
 839530
Total rural population in maine (estimation v3):
 819097
  • This command calculates the rural residents using three different approaches. 2022 rural residents data in census tract level is not available in ACS tables; therefore, we'll have to estimate the data.
    • approach 1: 2022 state data in Digital Equity Act of 2021 includes 8 total covered populations in state level. We can use the ratio (population in census tract level/population in state level) to estimate rural residents. However, this approach is not reliable because it can assign data to urban area without rural residents.
    • approach 2: ACS P2(Urban and Rural) 2020 Congressional District Summary File includes urban and rural populations in census tract level and block level. Based on the analysis, each block only includes rural or urban population. Therefore, we can assign rural/urban label to each block. Then we estimate 2022 population in block level and group them into census tract. If a block has rural label, we consider all residents in this block as rural residents.
    • approach 3: 2019 census tract data in Digital Equity Act of 2021 includes the percentage of rural residents in each census tract. We can use this percentage to estimate 2022 rural residents in each census tract.

To calculate SE scores and save the result into a .csv file, the following command is executed.

make SE
  • The output will be the top 5 areas(census tracts) in each type of covered populations. The result will be discussed in [results.md].
make SE_map

To combine SE and geo data and make maps by geo id. It will output a file named SE_normed_tract.png.

DDI (INFA + SE)

make DDI

In this step, we calculate the DDI of the each tract, delete the duplicates and scale DDI score to range 0 to 100. It will output a file named DDI.csv.

make DDI_map

This command will output the DDI map of Maine. The result file is named DDI_scaled_tract.png.

Analysis

To check the result of the analysis, the following commanded is executed.

make SE_analysis

Visualization

INFA MAP: 01 This is a test map for using code and visualize data, not final version. Since we found several different equations and data. From the fig of INFA, we can find that some tracts don't have data. In the middle, there are eight tracts have good performence. The worst tracts are in the north and east.

SE MAP: 02 In the SE results, the south parts where are close to Portland and have more population have good performence. They have more younger and well-educated people. Middle close to south tracts where are close to the sea have better performence than in the west. In the east tracts, the performence is not very well. The worst part is the north tracts close to the west.

Challenge

The major challenge is the data collection step.

  • Incarcerated individuals

    • The data cannot be find directly in the American Community Survey. In the census.gov, after choosing Population and People -> Counts, Estimates, and Projections -> Group Quarters Population -> B26103 Group Quarters Type(3 types), we can get Adult correctional facilities which is 3350. It's a estimated data and we cannot get the data of each tract.
    • In the National Institute Corrections, we can get the corrections state statistics information of Maine in 2021.
    • In the National Institute Correction, Prison Population: 1,473, Jail Population: 1,411, Parole Population: 17, Probation Population: 5,303.
    • The data in NIC and census.gov excludes Federal facilities.
    • The number of incarcerated people in Maine and each tract is estimated.
  • Rural

    • Rural residents data is incomplete because rural residents data (2018-2022) can not be found. H2 Urban and Rural Table might be added but it only contains data in 2020.

References

  • Because we used Min-Max normalization, the larget value of SE becomes 100.

About

project-spring24-hueyyt created by GitHub Classroom

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages