Exercice 1

L'idée était de vous faire mettre en forme un jeu de données vous-même, avant d'effectuer une ACP. En effet l'ACP directe n'est pas possible car il y a trop de lignes incomplètes, des colonnes a priori peu pertinentes, et des données de type séries temporelles (que l'on ramènera à une seule valeur).

# Import du jeu de données brut
data_orig <- read.csv("https://raw.githubusercontent.com/owid/covid-19-data/master/public/data/owid-covid-data.csv")

# Affichage d'un aperçu du jeu de données ; fonctions head() et summary() = bons réflexes :)
data <- data_orig
dim(data)
head(data)

1. 344917
2. 67

A data.frame: 6 × 67

	iso_code	continent	location	date	total_cases	new_cases	new_cases_smoothed	total_deaths	new_deaths	new_deaths_smoothed	⋯	male_smokers	handwashing_facilities	hospital_beds_per_thousand	life_expectancy	human_development_index	population	excess_mortality_cumulative_absolute	excess_mortality_cumulative	excess_mortality	excess_mortality_cumulative_per_million
	<chr>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	⋯	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
1	AFG	Asia	Afghanistan	2020-01-03	NA	0	NA	NA	0	NA	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA
2	AFG	Asia	Afghanistan	2020-01-04	NA	0	NA	NA	0	NA	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA
3	AFG	Asia	Afghanistan	2020-01-05	NA	0	NA	NA	0	NA	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA
4	AFG	Asia	Afghanistan	2020-01-06	NA	0	NA	NA	0	NA	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA
5	AFG	Asia	Afghanistan	2020-01-07	NA	0	NA	NA	0	NA	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA
6	AFG	Asia	Afghanistan	2020-01-08	NA	0	0	NA	0	0	⋯	NA	37.746	0.5	64.83	0.511	41128772	NA	NA	NA	NA

# On se limite pour l'instant à des statistiques globales prises à un jour fixé :
dmax <- '2021-12-31' #max(data$date)
data <- data[data$date == dmax,] #filtre sur les lignes, donc *avant* la virgule
dim(data)
summary(data)

1. 254
2. 67

   iso_code          continent           location             date          
 Length:254         Length:254         Length:254         Length:254        
 Class :character   Class :character   Class :character   Class :character  
 Mode  :character   Mode  :character   Mode  :character   Mode  :character  
                                                                            
                                                                            
                                                                            
                                                                            
  total_cases          new_cases       new_cases_smoothed   total_deaths    
 Min.   :        1   Min.   :      0   Min.   :      0.0   Min.   :      1  
 1st Qu.:    18326   1st Qu.:      0   1st Qu.:     37.0   1st Qu.:    281  
 Median :   148839   Median :    295   Median :    296.8   Median :   2468  
 Mean   :  5085552   Mean   :  22919   Mean   :  19680.6   Mean   : 101373  
 3rd Qu.:   806066   3rd Qu.:   1768   3rd Qu.:   1848.0   3rd Qu.:  16624  
 Max.   :285446097   Max.   :1349829   Max.   :1122946.0   Max.   :5474098  
 NA's   :19          NA's   :8         NA's   :8           NA's   :29       
   new_deaths      new_deaths_smoothed total_cases_per_million
 Min.   :   0.00   Min.   :   0.000    Min.   :     8.99      
 1st Qu.:   0.00   1st Qu.:   0.000    1st Qu.:  8437.41      
 Median :   0.00   Median :   1.214    Median : 57981.15      
 Mean   : 101.60   Mean   : 112.105    Mean   : 69265.17      
 3rd Qu.:   9.75   3rd Qu.:  13.607    3rd Qu.:108410.14      
 Max.   :5965.00   Max.   :6481.714    Max.   :289593.33      
 NA's   :8         NA's   :8           NA's   :19             
 new_cases_per_million new_cases_smoothed_per_million total_deaths_per_million
 Min.   :   0.00       Min.   :   0.000               Min.   :   1.086        
 1st Qu.:   0.00       1st Qu.:   7.484               1st Qu.: 139.656        
 Median :  48.22       Median :  83.628               Median : 720.977        
 Mean   : 543.64       Mean   : 370.260               Mean   :1020.674        
 3rd Qu.: 358.59       3rd Qu.: 426.848               3rd Qu.:1658.001        
 Max.   :6861.15       Max.   :3543.851               Max.   :5949.676        
 NA's   :8             NA's   :8                      NA's   :29              
 new_deaths_per_million new_deaths_smoothed_per_million reproduction_rate
 Min.   : 0.0000        Min.   : 0.0000                 Min.   :0.020    
 1st Qu.: 0.0000        1st Qu.: 0.0000                 1st Qu.:0.990    
 Median : 0.0000        Median : 0.1995                 Median :1.250    
 Mean   : 1.2384        Mean   : 1.3266                 Mean   :1.307    
 3rd Qu.: 0.9377        3rd Qu.: 1.4287                 3rd Qu.:1.570    
 Max.   :29.6820        Max.   :16.5200                 Max.   :4.220    
 NA's   :8              NA's   :8                       NA's   :69       
  icu_patients     icu_patients_per_million hosp_patients  
 Min.   :   19.0   Min.   : 0.579           Min.   :   67  
 1st Qu.:   75.5   1st Qu.:10.996           1st Qu.:  575  
 Median :  317.0   Median :17.341           Median : 1599  
 Mean   : 1129.5   Mean   :25.830           Mean   : 6510  
 3rd Qu.:  774.2   3rd Qu.:37.632           3rd Qu.: 3284  
 Max.   :18382.0   Max.   :84.912           Max.   :93776  
 NA's   :220       NA's   :220              NA's   :220    
 hosp_patients_per_million weekly_icu_admissions
 Min.   : 50.39            Min.   :   5.0       
 1st Qu.: 96.10            1st Qu.: 162.8       
 Median :157.06            Median : 523.0       
 Mean   :177.82            Mean   : 730.1       
 3rd Qu.:216.94            3rd Qu.:1130.8       
 Max.   :500.19            Max.   :1980.0       
 NA's   :220               NA's   :246          
 weekly_icu_admissions_per_million weekly_hosp_admissions
 Min.   : 0.529                    Min.   :  230         
 1st Qu.: 9.080                    1st Qu.:  894         
 Median :15.690                    Median : 2434         
 Mean   :15.395                    Mean   : 8879         
 3rd Qu.:22.003                    3rd Qu.: 8993         
 Max.   :29.198                    Max.   :92322         
 NA's   :246                       NA's   :235           
 weekly_hosp_admissions_per_million  total_tests          new_tests      
 Min.   : 30.10                     Min.   :    55939   Min.   :    162  
 1st Qu.: 74.18                     1st Qu.:  2056989   1st Qu.:   5700  
 Median :121.31                     Median :  7072918   Median :  18268  
 Mean   :133.89                     Mean   : 38588425   Mean   : 141706  
 3rd Qu.:203.27                     3rd Qu.: 28473905   3rd Qu.:  71131  
 Max.   :272.91                     Max.   :726359152   Max.   :2017702  
 NA's   :235                        NA's   :156         NA's   :161      
 total_tests_per_thousand new_tests_per_thousand new_tests_smoothed
 Min.   :   11.45         Min.   :  0.097        Min.   :      13  
 1st Qu.:  273.97         1st Qu.:  0.810        1st Qu.:    3442  
 Median :  954.34         Median :  2.616        Median :   11916  
 Mean   : 1876.38         Mean   :  7.742        Mean   :  201231  
 3rd Qu.: 2092.83         3rd Qu.:  5.987        3rd Qu.:   49380  
 Max.   :21654.04         Max.   :189.146        Max.   :14769984  
 NA's   :156              NA's   :161            NA's   :118       
 new_tests_smoothed_per_thousand positive_rate    tests_per_case   
 Min.   :  0.008                 Min.   :0.0000   Min.   :    1.5  
 1st Qu.:  0.477                 1st Qu.:0.0443   1st Qu.:    4.2  
 Median :  1.749                 Median :0.0970   Median :   10.3  
 Mean   :  5.240                 Mean   :0.1394   Mean   :  696.6  
 3rd Qu.:  4.349                 3rd Qu.:0.2386   3rd Qu.:   22.6  
 Max.   :110.367                 Max.   :0.6703   Max.   :65979.5  
 NA's   :118                     NA's   :127      NA's   :127      
 tests_units        total_vaccinations  people_vaccinated  
 Length:254         Min.   :1.050e+05   Min.   :5.568e+04  
 Class :character   1st Qu.:4.256e+06   1st Qu.:2.258e+06  
 Mode  :character   Median :1.549e+07   Median :8.206e+06  
                    Mean   :3.647e+08   Mean   :1.795e+08  
                    3rd Qu.:9.113e+07   3rd Qu.:4.509e+07  
                    Max.   :9.178e+09   Max.   :4.558e+09  
                    NA's   :155         NA's   :162        
 people_fully_vaccinated total_boosters      new_vaccinations  
 Min.   :4.931e+04       Min.   :     5245   Min.   :      45  
 1st Qu.:1.832e+06       1st Qu.:   834874   1st Qu.:    4035  
 Median :6.659e+06       Median :  2729817   Median :   28059  
 Mean   :1.506e+08       Mean   : 28731545   Mean   : 1611225  
 3rd Qu.:3.982e+07       3rd Qu.:  9342736   3rd Qu.:  186534  
 Max.   :3.879e+09       Max.   :536089356   Max.   :35457928  
 NA's   :161             NA's   :179         NA's   :169       
 new_vaccinations_smoothed total_vaccinations_per_hundred
 Min.   :       0          Min.   :  6.79                
 1st Qu.:     987          1st Qu.: 96.15                
 Median :   10899          Median :148.62                
 Mean   :  629396          Mean   :133.91                
 3rd Qu.:   69828          3rd Qu.:180.89                
 Max.   :35763420          Max.   :275.36                
 NA's   :23                NA's   :155                   
 people_vaccinated_per_hundred people_fully_vaccinated_per_hundred
 Min.   : 4.72                 Min.   : 2.05                      
 1st Qu.:50.73                 1st Qu.:43.51                      
 Median :69.69                 Median :63.24                      
 Mean   :62.04                 Mean   :55.63                      
 3rd Qu.:77.56                 3rd Qu.:71.47                      
 Max.   :93.19                 Max.   :89.24                      
 NA's   :162                   NA's   :161                        
 total_boosters_per_hundred new_vaccinations_smoothed_per_million
 Min.   : 0.020             Min.   :    0                        
 1st Qu.: 6.915             1st Qu.:  750                        
 Median :19.900             Median : 2253                        
 Mean   :22.071             Mean   : 2995                        
 3rd Qu.:33.240             3rd Qu.: 4310                        
 Max.   :57.400             Max.   :27586                        
 NA's   :179                NA's   :23                           
 new_people_vaccinated_smoothed new_people_vaccinated_smoothed_per_hundred
 Min.   :      0                Min.   :0.00000                           
 1st Qu.:    199                1st Qu.:0.01500                           
 Median :   2002                Median :0.03300                           
 Mean   : 121963                Mean   :0.06284                           
 3rd Qu.:  16102                3rd Qu.:0.06900                           
 Max.   :6989791                Max.   :0.50500                           
 NA's   :23                     NA's   :23                                
 stringency_index population_density    median_age    aged_65_older   
 Min.   : 6.48    Min.   :    0.137   Min.   :15.10   Min.   : 1.144  
 1st Qu.:35.19    1st Qu.:   38.612   1st Qu.:22.27   1st Qu.: 3.526  
 Median :43.95    Median :   90.672   Median :29.80   Median : 6.378  
 Mean   :44.51    Mean   :  444.976   Mean   :30.53   Mean   : 8.703  
 3rd Qu.:53.80    3rd Qu.:  225.097   3rd Qu.:38.80   3rd Qu.:13.928  
 Max.   :85.08    Max.   :20546.766   Max.   :48.20   Max.   :27.049  
 NA's   :73       NA's   :39          NA's   :54      NA's   :61      
 aged_70_older    gdp_per_capita     extreme_poverty  cardiovasc_death_rate
 Min.   : 0.526   Min.   :   661.2   Min.   : 0.100   Min.   : 79.37       
 1st Qu.: 2.099   1st Qu.:  4126.5   1st Qu.: 0.625   1st Qu.:174.59       
 Median : 3.893   Median : 12595.3   Median : 2.500   Median :245.06       
 Mean   : 5.499   Mean   : 19173.6   Mean   :13.848   Mean   :264.27       
 3rd Qu.: 8.638   3rd Qu.: 27341.8   3rd Qu.:21.350   3rd Qu.:331.93       
 Max.   :18.493   Max.   :116935.6   Max.   :77.600   Max.   :724.42       
 NA's   :56       NA's   :58         NA's   :128      NA's   :58           
 diabetes_prevalence female_smokers   male_smokers   handwashing_facilities
 Min.   : 0.990      Min.   : 0.10   Min.   : 7.70   Min.   :  1.188       
 1st Qu.: 5.378      1st Qu.: 1.90   1st Qu.:22.60   1st Qu.: 20.482       
 Median : 7.205      Median : 6.30   Median :33.10   Median : 49.691       
 Mean   : 8.561      Mean   :10.79   Mean   :32.91   Mean   : 50.789       
 3rd Qu.:10.770      3rd Qu.:19.20   3rd Qu.:41.30   3rd Qu.: 82.687       
 Max.   :30.530      Max.   :44.00   Max.   :78.10   Max.   :100.000       
 NA's   :48          NA's   :107     NA's   :109     NA's   :158           
 hospital_beds_per_thousand life_expectancy human_development_index
 Min.   : 0.100             Min.   :53.28   Min.   :0.3940         
 1st Qu.: 1.300             1st Qu.:69.59   1st Qu.:0.6030         
 Median : 2.500             Median :75.05   Median :0.7400         
 Mean   : 3.097             Mean   :73.74   Mean   :0.7225         
 3rd Qu.: 4.200             3rd Qu.:79.46   3rd Qu.:0.8287         
 Max.   :13.800             Max.   :86.75   Max.   :0.9570         
 NA's   :81                 NA's   :21      NA's   :64             
   population        excess_mortality_cumulative_absolute
 Min.   :4.700e+01   Min.   : -13147.4                   
 1st Qu.:4.677e+05   1st Qu.:    205.4                   
 Median :5.763e+06   Median :   4737.8                   
 Mean   :1.275e+08   Mean   :  52703.1                   
 3rd Qu.:2.827e+07   3rd Qu.:  22096.6                   
 Max.   :7.975e+09   Max.   :1076833.1                   
                     NA's   :189                         
 excess_mortality_cumulative excess_mortality
 Min.   :-6.47               Min.   :-24.71  
 1st Qu.: 6.30               1st Qu.:  3.85  
 Median :19.24               Median : 13.91  
 Mean   :17.43               Mean   : 18.83  
 3rd Qu.:25.67               3rd Qu.: 32.22  
 Max.   :50.05               Max.   : 76.05  
 NA's   :189                 NA's   :189     
 excess_mortality_cumulative_per_million
 Min.   :-1034.9                        
 1st Qu.:  782.7                        
 Median : 1773.4                        
 Mean   : 2213.6                        
 3rd Qu.: 3305.7                        
 Max.   : 7421.2                        
 NA's   :189                            

# Le jeu de données comprend quelques "lignes résumé" (continents, catégories de revenus...).
# En observant un peu, on remarque que leur code ISO démarre par "OWID_" :
owid_lines <- startsWith(data$iso_code, "OWID_")
data[owid_lines,]

A data.frame: 18 × 67

	iso_code	continent	location	date	total_cases	new_cases	new_cases_smoothed	total_deaths	new_deaths	new_deaths_smoothed	⋯	male_smokers	handwashing_facilities	hospital_beds_per_thousand	life_expectancy	human_development_index	population	excess_mortality_cumulative_absolute	excess_mortality_cumulative	excess_mortality	excess_mortality_cumulative_per_million
	<chr>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	⋯	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
2093	OWID_AFR		Africa	2021-12-31	9850360	58952	44044.714	228878	295	211.571	⋯	NA	NA	NA	NA	NA	1426736614	NA	NA	NA	NA
17104	OWID_ASI		Asia	2021-12-31	84647174	114105	85279.143	1256618	1055	1107.429	⋯	NA	NA	NA	NA	NA	4721383370	NA	NA	NA	NA
87984	OWID_ENG	Europe	England	2021-12-31	NA	NA	NA	NA	NA	NA	⋯	NA	NA	NA	NA	NA	56550000	NA	NA	NA	NA
96168	OWID_EUR		Europe	2021-12-31	86422548	565941	569201.000	1566172	2308	3111.857	⋯	NA	NA	NA	NA	NA	744807803	NA	NA	NA	NA
97538	OWID_EUN		European Union	2021-12-31	53608595	241084	351451.000	915208	1139	1654.286	⋯	NA	NA	NA	NA	NA	450146793	NA	NA	NA	NA
131649	OWID_HIC		High income	2021-12-31	134132754	1084099	912211.429	2042081	3297	3470.714	⋯	NA	NA	NA	NA	NA	1250514600	NA	NA	NA	NA
158897	OWID_KOS	Europe	Kosovo	2021-12-31	161399	53	20.571	2980	0	0.143	⋯	NA	NA	NA	NA	NA	1782115	6396.4	32.62	13.45	3848.565
173907	OWID_LIC		Low income	2021-12-31	1805520	23632	16829.571	42106	38	47.714	⋯	NA	NA	NA	NA	NA	737604900	NA	NA	NA	NA
175271	OWID_LMC		Lower middle income	2021-12-31	65603392	70197	55666.714	1187053	831	1072.143	⋯	NA	NA	NA	NA	NA	3432097300	NA	NA	NA	NA
221102	OWID_NAM		North America	2021-12-31	64102325	542800	357089.571	1224887	2012	1750.286	⋯	NA	NA	NA	NA	NA	600323657	NA	NA	NA	NA
224823	OWID_CYN	Asia	Northern Cyprus	2021-12-31	NA	NA	NA	NA	NA	NA	⋯	NA	NA	NA	NA	NA	382836	NA	NA	NA	NA
225824	OWID_NIR	Europe	Northern Ireland	2021-12-31	NA	NA	NA	NA	NA	NA	⋯	NA	NA	NA	NA	NA	1896000	NA	NA	NA	NA
229914	OWID_OCE		Oceania	2021-12-31	548908	670	14575.286	4989	14	13.714	⋯	NA	NA	NA	NA	NA	45038860	NA	NA	NA	NA
270779	OWID_SCT	Europe	Scotland	2021-12-31	NA	NA	NA	NA	NA	NA	⋯	NA	NA	NA	NA	NA	5466000	NA	NA	NA	NA
287145	OWID_SAM		South America	2021-12-31	39873056	67361	52751.429	1192550	281	286.857	⋯	NA	NA	NA	NA	NA	436816679	NA	NA	NA	NA
328061	OWID_UMC		Upper middle income	2021-12-31	83617794	169473	136426.714	2200159	1793	1887.857	⋯	NA	NA	NA	NA	NA	2525921300	NA	NA	NA	NA
337532	OWID_WLS	Europe	Wales	2021-12-31	NA	NA	NA	NA	NA	NA	⋯	NA	NA	NA	NA	NA	3170000	NA	NA	NA	NA
340184	OWID_WRL		World	2021-12-31	285446097	1349829	1122946.000	5474098	5965	6481.714	⋯	34.635	60.13	2.705	72.58	0.737	7975105024	NA	NA	NA	NA

# Toutes ces lignes spéciales sont a priori à supprimer, soit car elles correspondent à des aggrégations à retrouver plus tard,
# soit car elles concernent des pays avec très peu de données, à l'exception semble-t-il du Kosovo. Gardons seulement cette ligne.
kosovo_index <- data$iso_code == "OWID_KOS"
kosovo_line <- data[kosovo_index,]
kosovo_line[1] <- "KOS"
data <- rbind(data[!owid_lines,], kosovo_line)

# Pour supprimer les lignes avec valeurs manquantes, il y a plusieurs options.
# J'en indique ici 3, de la plus complexe à la plus simple :
data1 <- data[apply(data, 1, function(row) all(!is.na(row))),] #méthode 1
data2 <- data[complete.cases(data),]                           #méthode 2
data3 <- na.omit(data)                                         #méthode 3

# On remarque alors qu'il n'y a plus aucune ligne ==> il faut restreindre les colonnes
# (ou "deviner" les valeurs manquantes d'une manière ou d'une autre : voir le package missMDA)
# (ici on se contente de la version simple : pas de données manquantes en entrée).
nrow(data3)

0

# Que représentent les variables ?
colnames(data)

1. 'iso_code'
2. 'continent'
3. 'location'
4. 'date'
5. 'total_cases'
6. 'new_cases'
7. 'new_cases_smoothed'
8. 'total_deaths'
9. 'new_deaths'
10. 'new_deaths_smoothed'
11. 'total_cases_per_million'
12. 'new_cases_per_million'
13. 'new_cases_smoothed_per_million'
14. 'total_deaths_per_million'
15. 'new_deaths_per_million'
16. 'new_deaths_smoothed_per_million'
17. 'reproduction_rate'
18. 'icu_patients'
19. 'icu_patients_per_million'
20. 'hosp_patients'
21. 'hosp_patients_per_million'
22. 'weekly_icu_admissions'
23. 'weekly_icu_admissions_per_million'
24. 'weekly_hosp_admissions'
25. 'weekly_hosp_admissions_per_million'
26. 'total_tests'
27. 'new_tests'
28. 'total_tests_per_thousand'
29. 'new_tests_per_thousand'
30. 'new_tests_smoothed'
31. 'new_tests_smoothed_per_thousand'
32. 'positive_rate'
33. 'tests_per_case'
34. 'tests_units'
35. 'total_vaccinations'
36. 'people_vaccinated'
37. 'people_fully_vaccinated'
38. 'total_boosters'
39. 'new_vaccinations'
40. 'new_vaccinations_smoothed'
41. 'total_vaccinations_per_hundred'
42. 'people_vaccinated_per_hundred'
43. 'people_fully_vaccinated_per_hundred'
44. 'total_boosters_per_hundred'
45. 'new_vaccinations_smoothed_per_million'
46. 'new_people_vaccinated_smoothed'
47. 'new_people_vaccinated_smoothed_per_hundred'
48. 'stringency_index'
49. 'population_density'
50. 'median_age'
51. 'aged_65_older'
52. 'aged_70_older'
53. 'gdp_per_capita'
54. 'extreme_poverty'
55. 'cardiovasc_death_rate'
56. 'diabetes_prevalence'
57. 'female_smokers'
58. 'male_smokers'
59. 'handwashing_facilities'
60. 'hospital_beds_per_thousand'
61. 'life_expectancy'
62. 'human_development_index'
63. 'population'
64. 'excess_mortality_cumulative_absolute'
65. 'excess_mortality_cumulative'
66. 'excess_mortality'
67. 'excess_mortality_cumulative_per_million'

# Variables "new_*" : instantané journalier d'un certain indicateur. On ne s'y intéressera pas ici (cf. plus bas).
# De même pour les variables "weekly_*" (indicateurs hebdomadaires, j'imagine). Reste :
selection <- colnames(data)[!startsWith(colnames(data), "new_") & !startsWith(colnames(data), "weekly_")]

# Colonnes avec +50% de valeurs renseignées
selection <- selection[ apply(data[,selection], 2, function(col) sum(!is.na(col)) > nrow(data)/2) ]
selection

1. 'iso_code'
2. 'continent'
3. 'location'
4. 'date'
5. 'total_cases'
6. 'total_deaths'
7. 'total_cases_per_million'
8. 'total_deaths_per_million'
9. 'reproduction_rate'
10. 'positive_rate'
11. 'tests_per_case'
12. 'tests_units'
13. 'stringency_index'
14. 'population_density'
15. 'median_age'
16. 'aged_65_older'
17. 'aged_70_older'
18. 'gdp_per_capita'
19. 'extreme_poverty'
20. 'cardiovasc_death_rate'
21. 'diabetes_prevalence'
22. 'female_smokers'
23. 'male_smokers'
24. 'hospital_beds_per_thousand'
25. 'life_expectancy'
26. 'human_development_index'
27. 'population'

On y voit enfin plus clair :

• iso_code : identifiant d'un pays, sur 3 lettres
• location : nom du pays
• continent, date : heu, continent, et date =)
• total_cases : nombre total de cas enregistrés jusqu'à dmax
• total_deaths : nombre total de décès enregistrés jusqu'à dmax
• total_cases_per_million : nombre relatif de cas totaux (par million)
• total_deaths_per_million : nombre relatif de décès (par million)
• tests_units : "Units used by the location to report its testing data" https://github.com/owid/covid-19-data/blob/master/public/data/README.md
  en fait cette colonne ne contient que ". " ==> inutilisable.
• population : nombre d'habitants
• population_density : densité de population (par kilomètre carré)
• median_age : âge médian, 50% des gens sont plus jeunes et 50% plus vieux
• aged_65_older : pourcentage de personnes dépassant 65 ans
• aged_70_older : pareil avec 70 ans
• gdp_per_capita : PIB par habitant
• extreme_poverty : pourcentage de la population sous le seuil d'extrême pauvreté
• cardiovasc_death_rate : "Death rate from cardiovascular disease in 2017 (annual number of deaths per 100,000 people)"
• diabetes_prevalence : Diabetes prevalence (% of population aged 20 to 79) in 2017
• female_smokers : pourcentage de fumeuses
• male_smokers : pourcentage de fumeurs
• hospital_beds_per_thousand : nombre de lits d'hôpital par tranche de 1000 habitants
• life_expectancy : espérance de vie
• human_development_index : indice de développement humain

data$tests_units #???

1. ''
2. 'tests performed'
3. ''
4. ''
5. 'tests performed'
6. ''
7. ''
8. ''
9. 'tests performed'
10. 'tests performed'
11. ''
12. 'tests performed'
13. 'tests performed'
14. 'tests performed'
15. 'tests performed'
16. 'units unclear'
17. 'tests performed'
18. ''
19. 'tests performed'
20. 'tests performed'
21. 'tests performed'
22. ''
23. ''
24. 'samples tested'
25. 'tests performed'
26. ''
27. 'tests performed'
28. 'tests performed'
29. 'tests performed'
30. ''
31. ''
32. 'tests performed'
33. ''
34. ''
35. 'tests performed'
36. ''
37. 'tests performed'
38. 'tests performed'
39. ''
40. ''
41. ''
42. 'tests performed'
43. 'tests performed'
44. 'tests performed'
45. ''
46. ''
47. ''
48. 'people tested'
49. 'tests performed'
50. 'people tested'
51. ''
52. 'tests performed'
53. 'tests performed'
54. 'tests performed'
55. ''
56. 'tests performed'
57. ''
58. ''
59. 'samples tested'
60. 'people tested'
61. ''
62. 'tests performed'
63. 'tests performed'
64. ''
65. 'tests performed'
66. ''
67. 'tests performed'
68. 'people tested'
69. ''
70. 'tests performed'
71. 'tests performed'
72. 'people tested'
73. ''
74. ''
75. 'tests performed'
76. 'tests performed'
77. 'tests performed'
78. 'tests performed'
79. 'tests performed'
80. ''
81. 'samples tested'
82. ''
83. ''
84. ''
85. 'tests performed'
86. 'people tested'
87. ''
88. ''
89. ''
90. 'tests performed'
91. ''
92. ''
93. 'tests performed'
94. 'tests performed'
95. 'tests performed'
96. 'samples tested'
97. 'people tested'
98. 'tests performed'
99. 'tests performed'
100. 'tests performed'
101. ''
102. 'people tested'
103. 'tests performed'
104. 'samples tested'
105. 'people tested'
106. ''
107. 'tests performed'
108. ''
109. 'tests performed'
110. ''
111. 'tests performed'
112. ''
113. 'tests performed'
114. 'tests performed'
115. ''
116. ''
117. ''
118. 'samples tested'
119. 'tests performed'
120. 'tests performed'
121. 'tests performed'
122. ''
123. 'tests performed'
124. 'tests performed'
125. 'people tested'
126. 'samples tested'
127. ''
128. 'tests performed'
129. ''
130. ''
131. 'tests performed'
132. ''
133. ''
134. 'people tested'
135. ''
136. 'tests performed'
137. ''
138. 'samples tested'
139. ''
140. ''
141. 'people tested'
142. 'tests performed'
143. 'samples tested'
144. 'tests performed'
145. ''
146. 'samples tested'
147. 'tests performed'
148. ''
149. 'tests performed'
150. ''
151. ''
152. 'tests performed'
153. ''
154. 'samples tested'
155. 'tests performed'
156. ''
157. 'people tested'
158. ''
159. 'tests performed'
160. ''
161. 'tests performed'
162. 'tests performed'
163. ''
164. 'tests performed'
165. 'tests performed'
166. 'people tested'
167. ''
168. 'tests performed'
169. 'tests performed'
170. 'tests performed'
171. 'tests performed'
172. ''
173. 'tests performed'
174. 'tests performed'
175. 'samples tested'
176. ''
177. ''
178. 'people tested'
179. ''
180. ''
181. ''
182. 'tests performed'
183. ''
184. ''
185. ''
186. 'tests performed'
187. 'tests performed'
188. 'people tested'
189. ''
190. ''
191. ''
192. ''
193. 'tests performed'
194. 'tests performed'
195. ''
196. ''
197. 'people tested'
198. 'people tested'
199. 'tests performed'
200. 'tests performed'
201. 'tests performed'
202. ''
203. 'tests performed'
204. 'tests performed'
205. 'tests performed'
206. ''
207. 'people tested'
208. ''
209. ''
210. 'tests performed'
211. 'tests performed'
212. 'tests performed'
213. ''
214. ''
215. 'people tested'
216. 'people tested'
217. 'tests performed'
218. ''
219. ''
220. ''
221. 'tests performed'
222. 'tests performed'
223. 'tests performed'
224. 'tests performed'
225. 'tests performed'
226. 'tests performed'
227. 'people tested'
228. ''
229. ''
230. ''
231. ''
232. 'people tested'
233. ''
234. ''
235. 'tests performed'
236. 'tests performed'
237. 'tests performed'

selection <- selection[selection != "tests_units"]
newData <- na.omit(data[,selection])
nrow(newData)

72

92 lignes est raisonnable (proche de 50% de la taille du jeu de données initial). Cependant, pour être cohérent il faut en plus choisir un type de variable : absolu, ou relatif ? Je préfère les indicateurs relatifs (*_per_million, *_density) :

selection <- selection[!selection %in% c("total_cases", "total_deaths", "population")]
newData <- na.omit(data[,selection])
nrow(newData) #92 encore

72

rownames(newData) <- newData$iso_code #pour l'affichage des individus
newData <- newData[,-c(1,4)] #suppression des colonnes "code ISO" et "date", désormais inutiles

Note : toute l'analyse jusqu'ici aurait pu se faire aussi facilement avec un autre langage, Python par exemple.
À partir d'ici cependant, le package R FactoMineR est très pratique (pas d'équivalent Python (?!))

# ...On est enfin prêt pour l'ACP !
library(FactoMineR)
res.pca <- PCA(newData, quali.sup=1:2, ncp=6, graph=FALSE)

options(repr.plot.width=15, repr.plot.height=10)
plotInd <- plot(res.pca, choix="ind", invisible="quali")
plotVar <- plot(res.pca, choix="var")
library(gridExtra)
grid.arrange(plotInd, plotVar, ncol=2)

Warning message:
“ggrepel: 5 unlabeled data points (too many overlaps). Consider increasing max.overlaps”

Bon, il semble qu'aucun individu ne se détache. En fait certaines lignes ont presque toutes leurs valeurs renseignées, et une fois complétées à l'aide de Google on trouve des individus extrêmes (Monaco, Singapour). EDIT : ici je n'ai pas complété le fichier - vous pouvez le faire.

extremes <- which(data$location %in% c("Monaco", "Singapore"))
data[extremes,selection]

A data.frame: 2 × 23

	iso_code	continent	location	date	total_cases_per_million	total_deaths_per_million	reproduction_rate	positive_rate	tests_per_case	stringency_index	⋯	aged_70_older	gdp_per_capita	extreme_poverty	cardiovasc_death_rate	diabetes_prevalence	female_smokers	male_smokers	hospital_beds_per_thousand	life_expectancy	human_development_index
	<chr>	<chr>	<chr>	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	⋯	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
197910	MCO	Europe	Monaco	2021-12-31	138280.67	1041.353	0.93	NA	NA	34.13	⋯	NA	NA	NA	NA	5.46	NA	NA	13.8	86.75	NA
277597	SGP	Asia	Singapore	2021-12-31	49566.07	146.886	1.14	NA	NA	42.77	⋯	7.049	85535.38	NA	92.243	10.99	5.2	28.3	2.4	83.62	0.938

Côté variables, aged_65_older et aged_70_older apparaissent très corrélées (en fait même confondues). C'est logique, on gardera donc seulement aged_70_older après vérification numérique :

cor(newData[,c("aged_65_older", "aged_70_older")])

A matrix: 2 × 2 of type dbl

	aged_65_older	aged_70_older
aged_65_older	1.0000000	0.9939191
aged_70_older	0.9939191	1.0000000

newData <- subset(newData, select=-aged_65_older)

res.pca <- PCA(newData, quali.sup=1:2, ncp=6, graph=FALSE)

plotInd <- plot(res.pca, choix="ind", invisible="quali", habillage=1)
plotVar <- plot(res.pca, choix="var")
library(gridExtra)
grid.arrange(plotInd, plotVar, ncol=2)

Warning message:
“ggrepel: 2 unlabeled data points (too many overlaps). Consider increasing max.overlaps”

Environ 65% de l'inertie expliquée dans ce premier plan ACP (51 + 14). Le cercle des corrélations oppose logiquement "richesse" à droite (HDI, PIB/hab), avec "pauvreté" à gauche. Sur le nuage des individus cela correspond grossièrement à l'opposition Europe occidentale / Afrique (quelques exceptions : Tunisie, Seychelles, ...).

Il est intéressant de constater que les indicateurs de richesse sont très corrélés positivement au nombre de morts par millions, lui-même anti-corrélé avec extreme_poverty : le COVID frapperait plutôt les riches ? Mais pourquoi donc, puisque le virus est partout ? Et bien un élément de réponse se trouve dans ce même plan ACP : aged_70_older => on y vit plus vieux, et dans une moindre mesure diabetes_prevalence => plus de cas de diabète (à vérifier numériquement).

cor(newData[,c("aged_70_older", "total_deaths_per_million", "human_development_index",
               "diabetes_prevalence", "extreme_poverty")])

A matrix: 5 × 5 of type dbl

	aged_70_older	total_deaths_per_million	human_development_index	diabetes_prevalence	extreme_poverty
aged_70_older	1.00000000	0.6116174	0.8203092	-0.02528401	-0.5541654
total_deaths_per_million	0.61161735	1.0000000	0.5002665	0.14617300	-0.4433506
human_development_index	0.82030920	0.5002665	1.0000000	0.17607414	-0.7396914
diabetes_prevalence	-0.02528401	0.1461730	0.1760741	1.00000000	-0.3462567
extreme_poverty	-0.55416536	-0.4433506	-0.7396914	-0.34625674	1.0000000

newData

A data.frame: 72 × 20

	continent	location	total_cases_per_million	total_deaths_per_million	reproduction_rate	positive_rate	tests_per_case	stringency_index	population_density	median_age	aged_70_older	gdp_per_capita	extreme_poverty	cardiovasc_death_rate	diabetes_prevalence	female_smokers	male_smokers	hospital_beds_per_thousand	life_expectancy	human_development_index
	<chr>	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
ALB	Europe	Albania	73495.999	1130.064	1.48	0.1110	9.0	46.30	104.871	38.0	8.643	11803.431	1.1	304.195	10.08	7.1	51.2	2.890	78.57	0.795
ARG	South America	Argentina	127015.620	2596.686	2.11	0.3190	3.1	36.52	16.177	31.9	7.441	18933.907	0.6	191.032	5.50	16.2	27.7	5.000	76.67	0.845
AUS	Oceania	Australia	14004.709	93.325	2.20	0.0745	13.4	47.16	3.202	37.9	10.129	44648.710	0.5	107.791	5.07	13.0	16.5	3.840	83.44	0.944
AUT	Europe	Austria	141452.257	1867.753	1.18	0.0042	239.8	44.16	106.749	44.4	13.748	45436.686	0.7	145.183	6.35	28.4	30.9	7.370	81.54	0.922
BGD	Asia	Bangladesh	9262.063	163.985	1.43	0.0219	45.7	29.63	1265.036	27.5	3.262	3523.984	14.8	298.003	8.38	1.0	44.7	0.800	72.59	0.632
BEL	Europe	Belgium	179883.824	2429.409	1.25	0.1750	5.7	33.89	375.564	41.8	12.849	42658.576	0.2	114.898	4.29	25.1	31.4	5.640	81.63	0.931
BIH	Europe	Bosnia and Herzegovina	89830.928	4152.737	1.36	0.1867	5.4	35.19	68.496	42.5	10.711	11713.895	0.2	329.635	10.08	30.2	47.7	3.500	77.40	0.780
BGR	Europe	Bulgaria	108210.980	4501.357	1.42	0.0838	11.9	45.82	65.180	44.7	13.272	18563.307	1.5	424.688	5.81	30.1	44.4	7.454	75.05	0.816
CAN	North America	Canada	54674.470	779.054	1.49	0.2292	4.4	68.16	4.037	41.4	10.797	44017.591	0.5	105.599	7.37	12.0	16.6	2.500	82.43	0.929
CHL	South America	Chile	92058.065	1994.314	1.19	0.0258	38.8	31.42	24.282	35.4	6.938	22767.037	1.3	127.993	8.46	34.2	41.5	2.110	80.18	0.851
CHN	Asia	China	92.624	3.997	1.31	0.0000	65979.5	79.17	147.674	38.7	5.929	15308.712	0.7	261.899	9.74	1.9	48.4	4.340	76.91	0.761
COL	South America	Colombia	99059.263	2503.488	2.01	0.1350	7.4	49.40	44.223	32.2	4.312	13254.949	4.5	124.240	7.44	4.7	13.5	1.710	77.29	0.767
CRI	North America	Costa Rica	110206.152	1419.462	1.05	0.0910	11.0	45.31	96.079	33.6	5.694	15524.995	1.3	137.973	8.78	6.4	17.4	1.130	80.28	0.810
HRV	Europe	Croatia	176082.986	3099.722	1.31	0.3788	2.6	38.05	73.726	44.0	13.053	22669.797	0.7	253.782	5.59	34.3	39.9	5.540	78.49	0.851
DNK	Europe	Denmark	133231.468	553.529	1.34	0.1133	8.8	31.52	136.520	42.3	12.325	46682.515	0.2	114.767	6.41	19.3	18.8	2.500	80.90	0.940
DOM	North America	Dominican Republic	37160.446	378.134	1.62	0.0994	10.1	47.08	222.873	27.6	4.419	14600.861	1.6	266.653	8.20	8.5	19.1	1.600	74.08	0.756
ECU	South America	Ecuador	30320.534	1870.396	1.26	0.2760	3.6	53.30	66.939	28.1	4.458	10581.936	3.6	140.448	5.55	2.0	12.3	1.500	77.01	0.759
SLV	North America	El Salvador	19212.981	603.340	0.65	0.0130	76.9	31.48	307.811	27.6	5.417	7292.458	2.2	167.295	8.87	2.5	18.8	1.300	73.32	0.673
EST	Europe	Estonia	171039.256	1378.516	1.27	0.1241	8.1	37.29	31.033	42.7	13.491	29481.252	0.5	255.569	4.02	24.5	39.3	4.690	78.74	0.892
ETH	Africa	Ethiopia	3367.185	56.136	1.42	0.4357	2.3	40.74	104.957	19.8	2.063	1729.927	26.7	182.634	7.47	0.4	8.5	0.300	66.60	0.485
GMB	Africa	Gambia	3758.322	126.756	0.53	0.0534	18.7	13.89	207.566	17.5	1.417	1561.767	10.1	331.430	1.91	0.7	31.2	1.100	62.05	0.496
GEO	Asia	Georgia	249638.058	3685.518	1.02	0.0601	16.6	50.00	65.032	38.7	10.244	9745.079	4.2	496.218	7.11	5.3	55.5	2.600	73.77	0.812
GHA	Africa	Ghana	4364.905	39.013	1.17	0.2351	4.3	51.27	126.719	21.1	1.948	4227.630	12.0	298.245	4.97	0.3	7.7	0.900	64.07	0.611
GRC	Europe	Greece	104473.849	1974.488	1.75	0.0581	17.2	74.28	83.479	45.3	14.524	24574.382	1.5	175.695	4.55	35.3	52.0	4.210	82.24	0.888
HUN	Europe	Hungary	126053.645	3931.454	1.00	0.1975	5.1	27.96	108.043	43.4	11.976	26777.561	0.5	278.296	7.55	26.8	34.8	7.020	76.88	0.854
IND	Asia	India	24583.308	339.465	2.10	0.0101	98.7	68.64	450.419	28.2	3.414	6426.674	21.2	282.280	10.39	1.9	20.6	0.530	69.66	0.645
IDN	Asia	Indonesia	15472.592	523.025	1.16	0.0012	830.5	66.69	145.725	29.3	3.053	11188.744	5.7	342.864	6.32	2.8	76.1	1.040	71.72	0.718
IRN	Asia	Iran	69934.029	1485.840	0.86	0.0181	55.2	54.63	49.831	32.4	3.182	19082.620	0.2	270.308	9.59	0.8	21.1	1.500	76.68	0.783
IRL	Europe	Ireland	149793.912	1211.202	1.53	0.4300	2.3	42.84	69.874	38.7	8.678	67335.293	0.2	126.459	3.28	23.0	25.7	2.960	82.30	0.955
ISR	Asia	Israel	146252.196	874.061	2.10	0.0280	35.7	48.91	402.606	30.6	7.359	33132.320	0.5	93.320	6.74	15.4	35.4	2.990	82.97	0.919
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
MAR	Africa	Morocco	25656.966	396.284	2.25	0.0777	12.9	65.27	80.080	29.6	4.209	7485.013	1.0	419.146	7.14	0.8	47.1	1.100	76.68	0.686
MOZ	Africa	Mozambique	5587.555	60.541	1.48	0.6374	1.6	37.04	37.728	17.7	1.870	1136.103	62.9	329.942	3.30	5.1	29.1	0.700	60.85	0.456
MMR	Asia	Myanmar	9797.725	355.634	0.87	0.0116	86.3	80.56	81.721	29.1	3.120	5591.597	6.4	202.104	4.61	6.3	35.2	0.900	67.13	0.583
NPL	Asia	Nepal	27119.361	379.539	1.22	0.0250	40.0	33.33	204.430	25.0	3.212	2442.804	15.0	260.797	7.26	9.5	37.8	0.300	70.78	0.602
NOR	Europe	Norway	72669.388	256.518	1.10	0.2700	3.7	51.85	14.462	39.7	10.813	64800.057	0.2	114.316	5.31	19.6	20.7	3.600	82.40	0.957
PAK	Asia	Pakistan	5490.774	122.638	1.44	0.0088	113.5	47.52	255.573	23.5	2.780	5034.708	4.0	423.031	8.35	2.8	36.7	0.600	67.27	0.557
PAN	North America	Panama	111383.433	1684.215	1.71	0.1076	9.3	42.34	55.133	29.7	5.030	22267.037	2.2	128.346	8.33	2.4	9.9	2.300	78.51	0.815
PRY	South America	Paraguay	68738.907	2451.648	1.19	0.0445	22.5	40.74	17.144	26.5	3.833	8827.010	1.7	199.128	8.27	5.0	21.6	1.300	74.25	0.728
PRT	Europe	Portugal	124888.605	1838.990	1.55	0.0731	13.7	48.15	112.371	46.2	14.924	27936.896	0.5	127.842	9.85	16.3	30.0	3.390	82.05	0.864
ROU	Europe	Romania	91927.269	2986.581	1.62	0.0275	36.4	56.48	85.129	43.0	11.690	23313.199	5.7	370.946	9.74	22.9	37.1	6.892	76.05	0.828
RUS	Europe	Russia	72557.126	2134.289	0.82	0.0517	19.3	54.17	8.823	39.6	9.393	24765.954	0.1	431.297	6.18	23.4	58.3	8.050	72.58	0.824
SVK	Europe	Slovakia	149152.071	2947.662	0.83	0.1070	9.3	56.34	113.128	41.2	9.167	30155.152	0.7	287.959	7.29	23.1	37.7	5.820	77.54	0.860
ZAF	Africa	South Africa	57543.972	1520.372	0.78	0.2607	3.8	44.44	46.754	27.3	3.053	12294.876	18.9	200.380	5.52	8.1	33.2	2.320	64.13	0.709
KOR	Asia	South Korea	12259.772	108.558	0.79	0.0236	42.3	46.39	527.967	43.4	8.622	35938.374	0.2	85.998	6.80	6.2	40.9	12.270	83.03	0.916
ESP	Europe	Spain	128265.632	1919.210	1.74	0.3080	3.2	43.44	93.105	45.5	13.799	34272.360	1.0	99.403	7.17	27.4	31.4	2.970	83.56	0.904
LKA	Asia	Sri Lanka	26898.175	686.098	0.94	0.0702	14.2	53.80	341.955	34.1	5.331	11669.077	0.7	197.093	10.68	0.3	27.0	3.600	76.98	0.782
SWE	Europe	Sweden	125711.264	1454.592	1.57	0.1220	8.2	41.43	24.718	41.0	13.433	46949.283	0.5	133.982	4.79	18.8	18.9	2.220	82.80	0.945
THA	Asia	Thailand	31011.538	302.635	1.09	0.0744	13.4	43.06	135.132	40.1	6.890	16277.671	0.1	109.861	7.04	1.9	38.8	2.100	77.15	0.777
TLS	Asia	Timor	14789.405	90.957	0.70	0.0023	429.1	40.58	87.176	18.0	1.897	6570.102	30.3	335.346	6.86	6.3	78.1	5.900	69.50	0.606
TGO	Africa	Togo	3408.749	28.027	1.64	0.2567	3.9	21.76	143.366	19.4	1.525	1429.813	49.2	280.033	6.15	0.9	14.2	0.700	61.04	0.515
TUN	Africa	Tunisia	58743.540	2068.935	1.65	0.0627	15.9	35.20	74.228	32.7	5.075	10849.297	2.0	318.991	8.52	1.1	65.8	2.300	76.70	0.740
TUR	Asia	Turkey	110635.410	962.067	1.39	0.0847	11.8	35.55	104.914	31.6	5.061	25129.341	0.2	171.285	12.13	14.1	41.1	2.810	77.69	0.820
UGA	Africa	Uganda	3019.518	69.778	1.69	0.1726	5.8	73.15	213.759	16.4	1.308	1697.707	41.6	213.333	2.50	3.4	16.7	0.500	63.37	0.544
UKR	Europe	Ukraine	90829.360	2330.704	0.97	0.1988	5.0	39.49	77.390	41.4	11.133	7894.393	0.1	539.849	7.11	13.5	47.4	8.800	72.06	0.779
GBR	Europe	United Kingdom	199109.996	2619.105	1.24	0.1001	10.0	44.06	272.898	40.8	12.527	39753.244	0.2	122.137	4.28	20.0	24.7	2.540	81.32	0.932
USA	North America	United States	158249.753	2421.163	1.64	0.2600	3.8	47.65	35.608	38.3	9.732	54225.446	1.2	151.089	10.79	19.1	24.6	2.770	78.86	0.926
URY	South America	Uruguay	119875.973	1802.036	1.57	0.0773	12.9	28.70	19.751	35.6	10.361	20551.409	0.1	160.708	6.93	14.0	19.9	2.800	77.91	0.817
VNM	Asia	Vietnam	17632.268	329.922	1.03	0.0926	10.8	69.44	308.127	32.6	4.718	6171.884	2.0	245.465	6.00	1.0	45.9	2.600	75.40	0.704
ZMB	Africa	Zambia	12448.652	186.335	1.55	0.3133	3.2	37.96	22.995	17.7	1.542	3689.251	57.5	234.499	3.94	3.1	24.7	2.000	63.89	0.584
ZWE	Africa	Zimbabwe	12973.101	306.179	0.81	0.2858	3.5	47.22	42.729	19.6	1.882	1899.775	21.4	307.846	1.82	1.6	30.7	1.700	61.49	0.571

La corrélation (resp. anti-corrélation) aged_70_older avec HDI (resp. extreme_poverty) et total_deaths est vérifiée. De même, on observe une légère corrélation positive (resp. négative) entre diabetes_prevalence et HDI (resp. extreme_poverty).

Ensuite, le taux de fumeuses semble très corrélé à l'âge médian : les femmes auraient plus tendance à fumer dans les pays où l'on vit plus vieux (donc en général plus riches). Ce n'est pas le cas de male_smokers : le taux de fumeurs n'indique quant à lui pas grand chose. De même, et plus étonnament, le taux de mortalité par maladies cardiovasculaires (infarctus j'imagine) ne paraît corrélé à rien - si ce n'est justement et assez logiquement, la proportion de fumeurs : "According to the American Heart Association, cardiovascular disease accounts for about 800,000 U.S. deaths every year,5 making it the leading cause of all deaths in the United States. Of those, nearly 20 percent are due to cigarette smoking." [https://www.fda.gov/tobacco-products/health-effects-tobacco-use/how-smoking-affects-heart-health#]

La coloration par continents montre une opposition haut/bas entre Europe de l'est et Europe de l'ouest + USA/Canada/Israel/Corée/Australie. Il semble y avoir relativement plus de fumeurs en Géorgie/Ukraine/Russie. les pays d'Amérique centrale et du sud sont plus bas, donc a priori moins touchés par les décès par infarctus et comportant moins de fumeurs. Il n'y a pas assez de pays d'Océanie pour en dire grand chose, et l'Asie est répartie un peu partout, montrant une grande inhomogénéité en comparaison aux autres continents.

Vérifions notre analyse en regardant de plus près quelques individus :

indivs_indices <- rownames(newData) %in% c("LUX", "UKR", "NER", "ECU")
newData[indivs_indices,c("location", "total_deaths_per_million", "aged_70_older", "male_smokers",
                         "cardiovasc_death_rate", "human_development_index")]

A data.frame: 3 × 6

	location	total_deaths_per_million	aged_70_older	male_smokers	cardiovasc_death_rate	human_development_index
	<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
ECU	Ecuador	1870.396	4.458	12.3	140.448	0.759
LUX	Luxembourg	980.542	9.842	26.0	128.275	0.916
UKR	Ukraine	2330.704	11.133	47.4	539.849	0.779

Luxembourg : population âgée, HDI élevé, 2x moins de fumeurs qu'en Ukraine mais 2x + qu'en Equateur.
Niger : population jeune, HDI bas, peu de fumeurs, très peu de morts du COVID.

Bref, passons au second plan ACP :

plotInd <- plot(res.pca, choix="ind", invisible="quali", habillage=1, axes=3:4)
plotVar <- plot(res.pca, choix="var", axes=3:4)
library(gridExtra)
grid.arrange(plotInd, plotVar, ncol=2)

Warning message:
“ggrepel: 12 unlabeled data points (too many overlaps). Consider increasing max.overlaps”

Peu d'inertie expliquée dans ce plan (à peine plus de 17%), mais une observation intéressante : anti-corrélation population_density et total_deaths_per_million ? À vérifier numériquement bien sûr car cette dernière flèche est loin du bord. Ce serait cependant cohérent : densément peuplé => contaminations plus faciles => plus de cas => plus de personnes très fragiles touchées => plus de morts.

On note aussi l'anti-corrélation entre diabetes_prevalence et extreme_poverty, déjà un peu observée dans le premier plan. Vérification numérique :

indivs_indices <- rownames(newData) %in% c("MLT", "EGY", "MNE", "MWI")
newData[indivs_indices,c("location", "total_deaths_per_million", "population_density",
                         "diabetes_prevalence", "extreme_poverty")]

A data.frame: 2 × 5

	location	total_deaths_per_million	population_density	diabetes_prevalence	extreme_poverty
	<chr>	<dbl>	<dbl>	<dbl>	<dbl>
MWI	Malawi	115.411	197.519	3.94	71.4
MLT	Malta	924.445	1454.037	8.83	0.2

Opposition Égypte / Malawi vérifiée sur l'axe diabète/pauvreté, ainsi que Malte/Montenegro sur l'axe morts_par_million/densité.