Data and code for the figures in "Disparities in mobile phone ownership reflect inequities in access to healthcare"
├── data
│ ├── data_fig2.rda
│ └── data_fig3AC.rda
└── R code
├── d_manage.R
└── figures.R
data_fig2.rda includes the data necessary to reproduce Figure 2. Loading it will make a list called data_fig2 in the environment.
This list includes two lists:
- clinic_time
- Ndest
Those two lists have two vectors:
- phone
- nphone
clinic_time includes the vectors of the travel time to healthcare from mobile phone owners (phone) and non-owners (nphone).
Ndest includes the vectors of the number of recent travel destinations from mobile phone owners (phone) and non-owners (nphone).
data_fig3AC.rda includes the data necessary to reproduce Figure 3. Loading it will make a list called data_fig3AC in the environment.
It includes two lists:
- data_fig3A
- data_fig3C
Those lists includes elements necessary to reproduce Figure 3.
data_fig3A includes two lists:
- imputedN
- smoothed_pm
imputedN includes the 15 distribution (counts) of imputed values for travel time to healthcare. It is available for mobile phone owners (phone) and non-owners (nphone), men (men) and women (women).
Those four sets of distributions have a similar structure: a data.frame with 16 variables.
- The first variable includes the values for clinic_time.
- N1 to N15 include the count for those values in the 15 imputed data sets.
smoothed_pm includes the 15 distribution (probability mass) of imputed values for travel time to healthcare. It is available for mobile phone owners (phone) and non-owners (nphone), men (men) and women (women).
Those four sets of distributions have a similar structure: a data.frame with 17 variables.
- The first variable includes the values for clinic_time.
- pm1 to pm15 include the smoothed probability masses for those values in the 15 imputed data sets.
- pm is the pooled smoothed probability mass. It is the one used for Figure 3A.
The data necessary to plot Figure 3A allow to generate the data for Figure 3B and 3D:
- Figure 3B plots weighted averages of the four pooled smoothed proability masses plotted in Figure 3A. The weights are avaialble in Figure 3.
- Figure 3D is the difference between the two distributions plotted in Figure 3B.
This list has two elements:
- mean_clinic_time
- imput_boot
It is a data.frame with the average time to healthcare for mobile phone owners and the general population. The code for this calculation is available in R code.
It includes two vectors:
- phone
- total
They include the distribution of the two average travel times to healthcare. They were estimated by bootstrap. The calculation method and the percentile confidence interval estimation are further detailed in R code.
This script includes three functions used in the data management of the raw data.
This script includes the code of all the figures in the manuscript and the supplementary material that were generated on R.
Figure 1 was created on illustrator so no code is available.
The required data for Figure 2 is available in data.
The required data for Figure 3 is available in data.
The required data for Figure 4 is available in a Table in the supplementary material.