Translation-invariant functional clustering on COVID-19 deaths adjusted on population risk factors

The codes in this allow to reproduce the simulations presented in Cheam et al. (2023) https://hal.science/hal-03952739/.

Installation

The packages icamix, denpro and regpro, need to be installed from the source (because they were removed from the CRAN repository).

R CMD INSTALL icamix_1.0.6.tar.gz

R CMD INSTALL denpro_0.9.2.tar.gz

R CMD INSTALL regpro_0.1.1.tar.gz

The package and code execution require the installation of the following external libraries:

package_list <- c("regpro",
                  "forcats",
                  "latex2exp",
                  "ggplot2",
                  "fda",
                  "mixtools",
                  "HDclassif",
                  "VGAM",
                  "rwavelet",
                  "BiocManager",
                  "dtw",
                  "NMF",
                  "sClust",
                  "geoR",
                  "zoo",
                  "stringr",
                  "FactoMineR",
                  "xtable",
                  "gridExtra",
                  "dplyr",
                  "MASS",
                  "VarSelLCM")

Install missing packages:

isinstall <- sapply(package_list, 
                    function(x) x %in% rownames(installed.packages()))
package_list[isinstall]
sapply(package_list[!isinstall], install.packages)

Also install Biobase:

BiocManager::install("Biobase")

The codes are based on the Clustfun package, companion of the paper Cheam et al. (2023). To install the package, execute the following command in a terminal:

R CMD INSTALL --build Clustfun_1.0.0.tar.gz

Reproduction of figures and tables

Below, all the information are gathered to reproduce the numerical results presented in the paper.

Section 2 Description of the data

The dataset (COVIDfull.rda) was built from the files COVID-19_LUT.csv and Policy.rds downloaded from https://github.com/CSSEGISandData/COVID-19_Unified-Dataset Badr et al. (2023). The data and scripts are stored in the folder builddata-August2021/.

Load Section2.R to reproduce Figure 1.

source("Section2.r")

Section 4.1 Investigating the strengths of the proposed approach

Load Section4_1plot.R to reproduce Figures 2–4. The results are stored in resultsSection4_1.rda and can be re-executed running Section4_1run.R.

source("Section4_1plot.r")

Section 4.2 Comparing the proposed approach with other methods used for COVID-19 studies

Load Section4_2plot.R to reproduce Figures 5–6. The results are stored in resultsSection4_2.rda and can be executed running Section4_2run.R.

source("Section4_2plot.r")

Section 4.3 Investigating the robustness of the proposed approach

Load Section4_3plot.R to reproduce Figures 7–8. The results are stored in resultsSection4_3.rda and can be executed running Section4_3run.R.

source("Section4_3plot.r")

Investigating geographical disparities for COVID-19

Section 5.1 Population risk factors

The results of this section can be reproduced by running Section5-1.R

source("Section5-1.r")

	1	2	3	4	5	6
PM2.5_PopWtd	-0.5664122	0.0212741	-0.4345034	0.0000121	0.1993832	0.0540271
NO2_PopWtd	0.7958363	0.3504402	0.3072890	0.0025897	0.0413344	0.6924301
WorldPop_Density	0.5089838	0.0000000	-0.0061600	0.9530119	-0.3229638	0.0014989
Diabetes	0.2816515	0.4229080	0.7646337	0.0000000	-0.4502616	0.0000053
Obesity	0.3336598	0.0466550	0.8527638	0.0000000	-0.1532284	0.1403605
Smoking	-0.2871324	0.1308322	-0.0830285	0.4262674	0.8082869	0.0000000
COPD	0.4248760	0.0081631	0.7307828	0.0000000	0.1981437	0.0555693
CVD	0.3014341	0.5895446	0.8457820	0.0000000	0.0678408	0.5158928
HIV	0.3708583	0.0000000	0.2717158	0.0080689	-0.6904543	0.0000000
Hypertension	0.1964048	0.7902244	0.8214016	0.0000000	0.3441937	0.0006817
WorldPop_65	0.0528298	0.0438759	0.2901739	0.0045538	0.8435328	0.0000000

Table 2

Section 5.2 Clustering of the regions

The results of this section can be reproduced by running Section5-2.R

source("Section5-2.r")

1	2	3	4	5	6	7	8	9	10
1.000	2.000	3.000	4.000	5.000	6.000	7.000	8.000	9.000	10.000
-1377.804	-1237.954	-1187.733	-1136.491	-1114.798	-1100.366	-1065.731	-1054.002	-1055.569	-1034.302

Table 3

1	2	3	4
338.9082	387.0266	585.7333	1350.0617
387.0266	371.4739	534.7584	1185.9924
585.7333	534.7584	549.9975	958.6426
1350.0617	1185.9924	958.6426	1084.5153

Table 4

1	2	3	4	5	6	7
0.1407480	860.0971	397.0239	958.8668	638.6910	1.0105156	0.4079440
0.3312190	1030.5490	414.8785	1219.6178	711.3828	1.1112846	0.3566133
0.2196141	1463.0474	364.7504	1736.1461	350.8012	1.2270311	0.2840139
0.3084189	2532.9712	961.5437	2253.9434	888.8210	0.9426568	0.3233362

Table 5

Section 5.3 Clusters analysis example: disparities and policy decisions

The results of this section can be reproduced by running Section5-3.R

source("Section5-3.r")

References

Badr, Hamada S., Benjamin F. Zaitchik, Gaige H. Kerr, Nhat-Lan H. Nguyen, Yen-Ting Chen, Patrick Hinson, Josh M. Colston, et al. 2023. “Unified Real-Time Environmental-Epidemiological Data for Multiscale Modeling of the COVID-19 Pandemic.” Scientific Data 10 (1): 367.

Cheam, Amay, Marc Fredette, Matthieu Marbac, and Fabien Navarro. 2023. “Translation-invariant functional clustering on COVID-19 deaths adjusted on population risk factors.” Journal of the Royal Statistical Society Series C: Applied Statistics 72 (2): 387–413.