Librairies

On utilise ici la librairie tidyverse et gtsummary

library(tidyverse)
library(gtsummary)

Import du jeu de données and_vertebrates

La documentation est ici : https://lter.github.io/lterdatasampler/index.html

library(lterdatasampler)

Il s’agit de mesures de tailles et de poids effectuées sur des invertébrés (Truites fardées côtière et Salamandres) dans une forêt de l’Oregon aux USA.
Dans cette section, nous utiliserons les variables du jeu de données and_vertebrates contenu dans le package lterdatasampler :
- section : type de forêt, CC= clear cut forest,OG = upstream old growth coniferous forest
- weight_g : poids en gramme
- species : nom de l’espèce capturé (Cutthroat trout ; Coastal giant salamander ; Cascade torrent salamander)

Décrire une variable continue

On s’intéresse au poids des invertébrés, la variable weight_g. On va calculer la moyenne, l’écart-type, le minimum, le premier quartile, la médiane, le troisième quartile, le maximum, le nombre d’observation, le nombre de données renseignées et le nombre de données manquantes.

Codes

Une commande par indicateur [base R]

mean(and_vertebrates$weight_g,na.rm = T) # Moyenne

## [1] 8.902859

sd(and_vertebrates$weight_g,na.rm = T) # Ecart-Type

## [1] 10.67628

min(and_vertebrates$weight_g,na.rm = T) # Min

## [1] 0.09

quantile(and_vertebrates$weight_g,probs=0.25,na.rm = T) # Q1

##  25% 
## 1.51

quantile(and_vertebrates$weight_g,probs=0.5,na.rm = T) # Médiane

##  50% 
## 6.05

quantile(and_vertebrates$weight_g,probs=0.75,na.rm = T) # Q3

##   75% 
## 11.66

max(and_vertebrates$weight_g,na.rm = T) # Max

## [1] 134.59

length(and_vertebrates$weight_g) # Nombre valeur

## [1] 32209

sum(!is.na(and_vertebrates$weight_g)) #Nombre NON manquant

## [1] 18941

sum(is.na(and_vertebrates$weight_g)) #Nombre manquant

## [1] 13268

Une commande par indicateur [Tidyverse]

and_vertebrates %>%
  summarise(
    moyenne=mean(weight_g,na.rm = T),
    `écart-type`=sd(weight_g,na.rm = T),
    minimum=min(weight_g,na.rm = T),
    q1=quantile(weight_g,na.rm = T,probs=0.25),
    mediane=quantile(weight_g,na.rm = T,probs=0.55),
    q3=quantile(weight_g,na.rm = T,probs=0.75),
    maximum=max(weight_g,na.rm = T),
    N=n(),
    n_obs=sum(!is.na(weight_g)),
    n_miss_obs=sum(is.na(weight_g))
  )

## # A tibble: 1 × 10
##   moyenne `écart-type` minimum    q1 mediane    q3 maximum     N n_obs
##     <dbl>        <dbl>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <int> <int>
## 1    8.90         10.7    0.09  1.51    6.92  11.7    135. 32209 18941
## # … with 1 more variable: n_miss_obs <int>

En une seule commande

summary(and_vertebrates$weight_g,na.rm = T) # Moyenne

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.090   1.510   6.050   8.903  11.660 134.590   13268

En joli

# Statistique descritptive de weight_g VERY BEAUTIFUL [librairie gtsummary]
library(gtsummary)
and_vertebrates %>%
  tbl_summary(
    include = weight_g,
    type = all_continuous() ~ "continuous2",
    statistic = all_continuous() ~ c("{N_nonmiss}",
                                     "{N_miss}",
                                     "{min}",
                                     "{p25}",
                                     "{median}",
                                     "{p75}",
                                     "{max}",
                                     "{mean}",
                                     "{sd}"),
    missing = "no"
  )

Characteristic	N = 32,209
weight_g
N	18,941
N missing	13,268
Minimum	0
25%	2
Median	6
75%	12
Maximum	135
Mean	9
SD	11

En graphique

ggplot(and_vertebrates) +
  aes(x="",y=weight_g) +
  geom_boxplot(width=0.2) +
  labs(title="Plot of SRF_R per treatment",x="",y="Weight")

Décrire une variable continue d’un sous-groupe de notre jeu de données

On souhaite maintenant calculer les même indicateurs, la moyenne, l’écartype, et autre indicateur de la variable weight_g mais uniquement pour les invertébrés capturés dans les forêt coupé à blanc ie pour lesquelles la valeur de la variable section est égale à CC pour Clear Cut forest.

Codes

Une commande par indicateur [base R]

mean(and_vertebrates$weight_g[and_vertebrates$section=="CC"],na.rm = T) # Moyenne

## [1] 9.508932

sd(and_vertebrates$weight_g[and_vertebrates$section=="CC"],na.rm = T) # Ecart-Type

## [1] 10.7977

min(and_vertebrates$weight_g[and_vertebrates$section=="CC"],na.rm = T) # Min

## [1] 0.09

quantile(and_vertebrates$weight_g[and_vertebrates$section=="CC"],probs=0.25,na.rm = T) # Q1

##   25% 
## 1.655

quantile(and_vertebrates$weight_g[and_vertebrates$section=="CC"],probs=0.5,na.rm = T) # Médiane

## 50% 
## 6.6

quantile(and_vertebrates$weight_g[and_vertebrates$section=="CC"],probs=0.75,na.rm = T) # Q3

##   75% 
## 12.55

max(and_vertebrates$weight_g[and_vertebrates$section=="CC"],na.rm = T) # Max

## [1] 102.99

length(and_vertebrates$weight_g[and_vertebrates$section=="CC"]) # Nombre valeur

## [1] 16490

sum(!is.na(and_vertebrates$weight_g[and_vertebrates$section=="CC"])) #Nombre NON manquant

## [1] 9831

sum(is.na(and_vertebrates$weight_g[and_vertebrates$section=="CC"])) #Nombre manquant

## [1] 6659

Une commande par indicateur [Tidyverse]

and_vertebrates %>%
  filter(section=="CC") %>%
  summarise(
    moyenne=mean(weight_g,na.rm = T),
    `écart-type`=sd(weight_g,na.rm = T),
    minimum=min(weight_g,na.rm = T),
    q1=quantile(weight_g,na.rm = T,probs=0.25),
    mediane=quantile(weight_g,na.rm = T,probs=0.55),
    q3=quantile(weight_g,na.rm = T,probs=0.75),
    maximum=max(weight_g,na.rm = T),
    N=n(),
    n_obs=sum(!is.na(weight_g)),
    n_miss_obs=sum(is.na(weight_g))
  )

## # A tibble: 1 × 10
##   moyenne `écart-type` minimum    q1 mediane    q3 maximum     N n_obs
##     <dbl>        <dbl>   <dbl> <dbl>   <dbl> <dbl>   <dbl> <int> <int>
## 1    9.51         10.8    0.09  1.65    7.54  12.6    103. 16490  9831
## # … with 1 more variable: n_miss_obs <int>

En une seule commande

summary(and_vertebrates$weight_g[and_vertebrates$section=="CC"],na.rm = T) # Moyenne

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.090   1.655   6.600   9.509  12.550 102.990    6659

En joli

librairie gtsummary

and_vertebrates %>% 
  filter(section=="CC") %>%
  tbl_summary(
    include = weight_g,
    type = all_continuous() ~ "continuous2",
    statistic = all_continuous() ~ c("{N_nonmiss}",
                                     "{N_miss}",
                                     "{min}",
                                     "{p25}",
                                     "{median}",
                                     "{p75}",
                                     "{max}",
                                     "{mean}",
                                     "{sd}"),
    missing = "no"
  )

Characteristic	N = 16,490
weight_g
N	9,831
N missing	6,659
Minimum	0
25%	2
Median	7
75%	13
Maximum	103
Mean	10
SD	11

En graphique

# Boxplot
and_vertebrates %>% 
  filter(section=="CC") %>%
  ggplot() +
  aes(x="",y=weight_g) +
  geom_boxplot(width=0.2) +
  labs(title="Plot of Weight for CC forest",x="",y="Weight")

Décrire une variable continue en fonction d’une variable catégorielle

On va maintenant décrire le poids des invertébrés en fonction de l’espèce capturé (Cutthroat trout ; Coastal giant salamander ; Cascade torrent salamander) en utilisant la variable species.

Codes

Une commande par indicateur [base R 1]

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = mean,  na.rm = TRUE)  # Moyenne

##                      Group.1        x
## 1 Cascade torrent salamander 1.082222
## 2   Coastal giant salamander 9.031720
## 3            Cutthroat trout 8.843599

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = sd,  na.rm = TRUE)  # Ecart-Type

##                      Group.1          x
## 1 Cascade torrent salamander  0.4727255
## 2   Coastal giant salamander 12.1137205
## 3            Cutthroat trout  9.8758610

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = min,  na.rm = TRUE) # Min

##                      Group.1    x
## 1 Cascade torrent salamander 0.50
## 2   Coastal giant salamander 0.29
## 3            Cutthroat trout 0.09

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = quantile, probs=0.25,  na.rm = TRUE)  # Q1

##                      Group.1   x
## 1 Cascade torrent salamander 0.8
## 2   Coastal giant salamander 2.9
## 3            Cutthroat trout 1.2

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = quantile, probs=0.55,  na.rm = TRUE)  # Médiane

##                      Group.1     x
## 1 Cascade torrent salamander 1.044
## 2   Coastal giant salamander 6.700
## 3            Cutthroat trout 7.050

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = quantile, probs=0.75,  na.rm = TRUE)  # Q3

##                      Group.1     x
## 1 Cascade torrent salamander  1.30
## 2   Coastal giant salamander 10.42
## 3            Cutthroat trout 12.90

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = max,  na.rm = TRUE) # Max

##                      Group.1      x
## 1 Cascade torrent salamander   1.95
## 2   Coastal giant salamander 134.59
## 3            Cutthroat trout 104.00

aggregate(x = and_vertebrates$weight_g, by = list(and_vertebrates$species), FUN = length) # Nombre valeur

##                      Group.1     x
## 1 Cascade torrent salamander    15
## 2   Coastal giant salamander 11758
## 3            Cutthroat trout 20433

table(and_vertebrates$species[!is.na(and_vertebrates$weight_g)]) #Nombre NON manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          9                       6338 
##            Cutthroat trout 
##                      12594

table(and_vertebrates$species[is.na(and_vertebrates$weight_g)]) #Nombre manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          6                       5420 
##            Cutthroat trout 
##                       7839

Une commande par indicateur [base R 2]

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = mean,  na.rm = TRUE)  # Moyenne

## Cascade torrent salamander   Coastal giant salamander 
##                   1.082222                   9.031720 
##            Cutthroat trout 
##                   8.843599

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = sd,  na.rm = TRUE)  # Ecart-Type

## Cascade torrent salamander   Coastal giant salamander 
##                  0.4727255                 12.1137205 
##            Cutthroat trout 
##                  9.8758610

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = min,  na.rm = TRUE) # Min

## Cascade torrent salamander   Coastal giant salamander 
##                       0.50                       0.29 
##            Cutthroat trout 
##                       0.09

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = quantile, probs=0.25,  na.rm = TRUE)  # Q1

## Cascade torrent salamander   Coastal giant salamander 
##                        0.8                        2.9 
##            Cutthroat trout 
##                        1.2

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = quantile, probs=0.55,  na.rm = TRUE)  # Médiane

## Cascade torrent salamander   Coastal giant salamander 
##                      1.044                      6.700 
##            Cutthroat trout 
##                      7.050

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = quantile, probs=0.75,  na.rm = TRUE)  # Q3

## Cascade torrent salamander   Coastal giant salamander 
##                       1.30                      10.42 
##            Cutthroat trout 
##                      12.90

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = max,  na.rm = TRUE) # Max

## Cascade torrent salamander   Coastal giant salamander 
##                       1.95                     134.59 
##            Cutthroat trout 
##                     104.00

tapply(and_vertebrates$weight_g, and_vertebrates$species, FUN = length) # Nombre valeur

## Cascade torrent salamander   Coastal giant salamander 
##                         15                      11758 
##            Cutthroat trout 
##                      20433

table(and_vertebrates$species[!is.na(and_vertebrates$weight_g)]) #Nombre NON manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          9                       6338 
##            Cutthroat trout 
##                      12594

table(and_vertebrates$species[is.na(and_vertebrates$weight_g)]) #Nombre manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          6                       5420 
##            Cutthroat trout 
##                       7839

Avec Tidyverse 1

and_vertebrates %>% group_by(species) %>% summarise(mean(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `mean(weight_g, na.rm = T)`
##   <chr>                                            <dbl>
## 1 Cascade torrent salamander                        1.08
## 2 Coastal giant salamander                          9.03
## 3 Cutthroat trout                                   8.84
## 4 <NA>                                            NaN

and_vertebrates %>% group_by(species) %>% summarise(sd(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `sd(weight_g, na.rm = T)`
##   <chr>                                          <dbl>
## 1 Cascade torrent salamander                     0.473
## 2 Coastal giant salamander                      12.1  
## 3 Cutthroat trout                                9.88 
## 4 <NA>                                          NA

and_vertebrates %>% group_by(species) %>% summarise(min(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `min(weight_g, na.rm = T)`
##   <chr>                                           <dbl>
## 1 Cascade torrent salamander                       0.5 
## 2 Coastal giant salamander                         0.29
## 3 Cutthroat trout                                  0.09
## 4 <NA>                                           Inf

and_vertebrates %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.25,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.25, na.rm = T)`
##   <chr>                                                              <dbl>
## 1 Cascade torrent salamander                                           0.8
## 2 Coastal giant salamander                                             2.9
## 3 Cutthroat trout                                                      1.2
## 4 <NA>                                                                NA

and_vertebrates %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.5,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.5, na.rm = T)`
##   <chr>                                                             <dbl>
## 1 Cascade torrent salamander                                         0.94
## 2 Coastal giant salamander                                           5.9 
## 3 Cutthroat trout                                                    6.15
## 4 <NA>                                                              NA

and_vertebrates %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.75,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.75, na.rm = T)`
##   <chr>                                                              <dbl>
## 1 Cascade torrent salamander                                           1.3
## 2 Coastal giant salamander                                            10.4
## 3 Cutthroat trout                                                     12.9
## 4 <NA>                                                                NA

and_vertebrates %>% group_by(species) %>% summarise(max(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `max(weight_g, na.rm = T)`
##   <chr>                                           <dbl>
## 1 Cascade torrent salamander                       1.95
## 2 Coastal giant salamander                       135.  
## 3 Cutthroat trout                                104   
## 4 <NA>                                          -Inf

and_vertebrates %>% group_by(species) %>% count()  # Nombre valeur

## # A tibble: 4 × 2
## # Groups:   species [4]
##   species                        n
##   <chr>                      <int>
## 1 Cascade torrent salamander    15
## 2 Coastal giant salamander   11758
## 3 Cutthroat trout            20433
## 4 <NA>                           3

and_vertebrates %>% group_by(species) %>% summarise(sum(!is.na(weight_g))) #Nombre NON manquant

## # A tibble: 4 × 2
##   species                    `sum(!is.na(weight_g))`
##   <chr>                                        <int>
## 1 Cascade torrent salamander                       9
## 2 Coastal giant salamander                      6338
## 3 Cutthroat trout                              12594
## 4 <NA>                                             0

and_vertebrates %>% group_by(species) %>% summarise(sum(is.na(weight_g))) #Nombre manquant

## # A tibble: 4 × 2
##   species                    `sum(is.na(weight_g))`
##   <chr>                                       <int>
## 1 Cascade torrent salamander                      6
## 2 Coastal giant salamander                     5420
## 3 Cutthroat trout                              7839
## 4 <NA>                                            3

Avec Tidyverse 2

and_vertebrates %>%
  group_by(species) %>% 
  summarise(mean=mean(weight_g,na.rm=T),
            sd=sd(weight_g,na.rm=T),
            min=min(weight_g,na.rm=T),
            q1=quantile(weight_g,probs=0.25,na.rm=T),
            mediane=quantile(weight_g,probs=0.5,na.rm=T),
            q3=quantile(weight_g,probs=0.75,na.rm=T),
            max=max(weight_g,na.rm=T),
            N=n(),
            n=sum(!is.na(weight_g)),
            missing=sum(is.na(weight_g)))

## # A tibble: 4 × 11
##   species     mean     sd    min    q1 mediane    q3     max     N     n missing
##   <chr>      <dbl>  <dbl>  <dbl> <dbl>   <dbl> <dbl>   <dbl> <int> <int>   <int>
## 1 Cascade …   1.08  0.473   0.5    0.8    0.94   1.3    1.95    15     9       6
## 2 Coastal …   9.03 12.1     0.29   2.9    5.9   10.4  135.   11758  6338    5420
## 3 Cutthroa…   8.84  9.88    0.09   1.2    6.15  12.9  104    20433 12594    7839
## 4 <NA>      NaN    NA     Inf     NA     NA     NA   -Inf        3     0       3

En une seule commande [base R]

tapply(and_vertebrates$weight_g, and_vertebrates$species, summary)

## $`Cascade torrent salamander`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.500   0.800   0.940   1.082   1.300   1.950       6 
## 
## $`Coastal giant salamander`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.290   2.900   5.900   9.032  10.420 134.590    5420 
## 
## $`Cutthroat trout`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.090   1.200   6.150   8.844  12.900 104.000    7839

En une seule commande [tidyverse]

and_vertebrates %>%  select(weight_g,species) %>%  split(.$species) %>% map(summary)

## $`Cascade torrent salamander`
##     weight_g       species         
##  Min.   :0.500   Length:15         
##  1st Qu.:0.800   Class :character  
##  Median :0.940   Mode  :character  
##  Mean   :1.082                     
##  3rd Qu.:1.300                     
##  Max.   :1.950                     
##  NA's   :6                         
## 
## $`Coastal giant salamander`
##     weight_g         species         
##  Min.   :  0.290   Length:11758      
##  1st Qu.:  2.900   Class :character  
##  Median :  5.900   Mode  :character  
##  Mean   :  9.032                     
##  3rd Qu.: 10.420                     
##  Max.   :134.590                     
##  NA's   :5420                        
## 
## $`Cutthroat trout`
##     weight_g         species         
##  Min.   :  0.090   Length:20433      
##  1st Qu.:  1.200   Class :character  
##  Median :  6.150   Mode  :character  
##  Mean   :  8.844                     
##  3rd Qu.: 12.900                     
##  Max.   :104.000                     
##  NA's   :7839

En beau

# [librairie gtsummary]
and_vertebrates %>% 
  tbl_summary(
    include = weight_g,
    by = species,
    type = all_continuous() ~ "continuous2",
    statistic = all_continuous() ~ c("{N_nonmiss}",
                                     "{N_miss}",
                                     "{min}",
                                     "{p25}",
                                     "{median}",
                                     "{p75}",
                                     "{max}",
                                     "{mean}",
                                     "{sd}"),
    missing = "no"
  ) %>%
  add_overall()

Characteristic	Overall, N = 32,206	Cascade torrent salamander, N = 15	Coastal giant salamander, N = 11,758	Cutthroat trout, N = 20,433
weight_g
N	18,941	9	6,338	12,594
N missing	13,265	6	5,420	7,839
Minimum	0	1	0	0
25%	2	1	3	1
Median	6	1	6	6
75%	12	1	10	13
Maximum	135	2	135	104
Mean	9	1	9	9
SD	11	0	12	10

En graphique

and_vertebrates %>% 
  drop_na(weight_g) %>% 
  ggplot() +
  aes(x=species, y=weight_g,color=species) +
  geom_boxplot() +
  labs(title="Plot of weight per species",x="species", y = "weight") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

and_vertebrates %>% 
  drop_na(weight_g) %>% 
  ggplot() +
  aes(x=species, y=weight_g,color=species) +
  geom_boxplot() +
  labs(title="Plot of weight per species",x="species", y = "weight") +
  coord_flip()

Décrire une variable continue en fonction d’une variable catégorielle d’un sous-groupe de notre jeu de données

On va maintenant décrire le poids des invertébrés en fonction de l’espèce capturé (Cutthroat trout ; Coastal giant salamander ; Cascade torrent salamander) en utilisant la variable species mais uniquement pour les invertébrés capturés dans les forêt coupé à blanc ie pour lesquelles la valeur de la variable section est égale à CC pour Clear Cut forest.

Codes

Une commande par indicateur [base R 1]

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = mean,  na.rm = TRUE)  # Moyenne

##                      Group.1        x
## 1 Cascade torrent salamander 1.318000
## 2   Coastal giant salamander 9.810634
## 3            Cutthroat trout 9.380571

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = sd,  na.rm = TRUE)  # Ecart-Type

##                      Group.1          x
## 1 Cascade torrent salamander  0.4718262
## 2   Coastal giant salamander 11.4887686
## 3            Cutthroat trout 10.4751060

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = min,  na.rm = TRUE) # Min

##                      Group.1    x
## 1 Cascade torrent salamander 0.80
## 2   Coastal giant salamander 0.29
## 3            Cutthroat trout 0.09

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = quantile, probs=0.25,  na.rm = TRUE)  # Q1

##                      Group.1    x
## 1 Cascade torrent salamander 0.94
## 2   Coastal giant salamander 3.40
## 3            Cutthroat trout 1.30

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = quantile, probs=0.55,  na.rm = TRUE)  # Médiane

##                      Group.1      x
## 1 Cascade torrent salamander 1.3600
## 2   Coastal giant salamander 7.6585
## 3            Cutthroat trout 7.4900

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = quantile, probs=0.75,  na.rm = TRUE)  # Q3

##                      Group.1    x
## 1 Cascade torrent salamander  1.6
## 2   Coastal giant salamander 11.5
## 3            Cutthroat trout 13.4

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = max,  na.rm = TRUE) # Max

##                      Group.1      x
## 1 Cascade torrent salamander   1.95
## 2   Coastal giant salamander 102.99
## 3            Cutthroat trout  77.78

aggregate(x = and_vertebrates$weight_g[and_vertebrates$section=="CC"], by = list(and_vertebrates$species[and_vertebrates$section=="CC"]), FUN = length) # Nombre valeur

##                      Group.1     x
## 1 Cascade torrent salamander     9
## 2   Coastal giant salamander  5406
## 3            Cutthroat trout 11073

table(and_vertebrates$species[and_vertebrates$section=="CC" & !is.na(and_vertebrates$weight_g)]) #Nombre NON manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          5                       3028 
##            Cutthroat trout 
##                       6798

table(and_vertebrates$species[and_vertebrates$section=="CC" & is.na(and_vertebrates$weight_g)]) #Nombre manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          4                       2378 
##            Cutthroat trout 
##                       4275

Une commande par indicateur [base R 2]

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = mean,  na.rm = TRUE)  # Moyenne

## Cascade torrent salamander   Coastal giant salamander 
##                   1.318000                   9.810634 
##            Cutthroat trout 
##                   9.380571

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = sd,  na.rm = TRUE)  # Ecart-Type

## Cascade torrent salamander   Coastal giant salamander 
##                  0.4718262                 11.4887686 
##            Cutthroat trout 
##                 10.4751060

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = min,  na.rm = TRUE) # Min

## Cascade torrent salamander   Coastal giant salamander 
##                       0.80                       0.29 
##            Cutthroat trout 
##                       0.09

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = quantile, probs=0.25,  na.rm = TRUE)  # Q1

## Cascade torrent salamander   Coastal giant salamander 
##                       0.94                       3.40 
##            Cutthroat trout 
##                       1.30

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = quantile, probs=0.55,  na.rm = TRUE)  # Médiane

## Cascade torrent salamander   Coastal giant salamander 
##                     1.3600                     7.6585 
##            Cutthroat trout 
##                     7.4900

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = quantile, probs=0.75,  na.rm = TRUE)  # Q3

## Cascade torrent salamander   Coastal giant salamander 
##                        1.6                       11.5 
##            Cutthroat trout 
##                       13.4

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = max,  na.rm = TRUE) # Max

## Cascade torrent salamander   Coastal giant salamander 
##                       1.95                     102.99 
##            Cutthroat trout 
##                      77.78

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], FUN = length) # Nombre valeur

## Cascade torrent salamander   Coastal giant salamander 
##                          9                       5406 
##            Cutthroat trout 
##                      11073

table(and_vertebrates$species[and_vertebrates$section=="CC" & !is.na(and_vertebrates$weight_g)]) #Nombre NON manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          5                       3028 
##            Cutthroat trout 
##                       6798

table(and_vertebrates$species[and_vertebrates$section=="CC" & is.na(and_vertebrates$weight_g)]) #Nombre manquant

## 
## Cascade torrent salamander   Coastal giant salamander 
##                          4                       2378 
##            Cutthroat trout 
##                       4275

Avec Tidyverse 1

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(mean(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `mean(weight_g, na.rm = T)`
##   <chr>                                            <dbl>
## 1 Cascade torrent salamander                        1.32
## 2 Coastal giant salamander                          9.81
## 3 Cutthroat trout                                   9.38
## 4 <NA>                                            NaN

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(sd(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `sd(weight_g, na.rm = T)`
##   <chr>                                          <dbl>
## 1 Cascade torrent salamander                     0.472
## 2 Coastal giant salamander                      11.5  
## 3 Cutthroat trout                               10.5  
## 4 <NA>                                          NA

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(min(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `min(weight_g, na.rm = T)`
##   <chr>                                           <dbl>
## 1 Cascade torrent salamander                       0.8 
## 2 Coastal giant salamander                         0.29
## 3 Cutthroat trout                                  0.09
## 4 <NA>                                           Inf

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.25,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.25, na.rm = T)`
##   <chr>                                                              <dbl>
## 1 Cascade torrent salamander                                          0.94
## 2 Coastal giant salamander                                            3.4 
## 3 Cutthroat trout                                                     1.3 
## 4 <NA>                                                               NA

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.5,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.5, na.rm = T)`
##   <chr>                                                             <dbl>
## 1 Cascade torrent salamander                                          1.3
## 2 Coastal giant salamander                                            6.7
## 3 Cutthroat trout                                                     6.6
## 4 <NA>                                                               NA

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(quantile(weight_g,probs=0.75,na.rm=T))

## # A tibble: 4 × 2
##   species                    `quantile(weight_g, probs = 0.75, na.rm = T)`
##   <chr>                                                              <dbl>
## 1 Cascade torrent salamander                                           1.6
## 2 Coastal giant salamander                                            11.5
## 3 Cutthroat trout                                                     13.4
## 4 <NA>                                                                NA

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(max(weight_g,na.rm=T))

## # A tibble: 4 × 2
##   species                    `max(weight_g, na.rm = T)`
##   <chr>                                           <dbl>
## 1 Cascade torrent salamander                       1.95
## 2 Coastal giant salamander                       103.  
## 3 Cutthroat trout                                 77.8 
## 4 <NA>                                          -Inf

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% count()  # Nombre valeur

## # A tibble: 4 × 2
## # Groups:   species [4]
##   species                        n
##   <chr>                      <int>
## 1 Cascade torrent salamander     9
## 2 Coastal giant salamander    5406
## 3 Cutthroat trout            11073
## 4 <NA>                           2

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(sum(!is.na(weight_g))) #Nombre NON manquant

## # A tibble: 4 × 2
##   species                    `sum(!is.na(weight_g))`
##   <chr>                                        <int>
## 1 Cascade torrent salamander                       5
## 2 Coastal giant salamander                      3028
## 3 Cutthroat trout                               6798
## 4 <NA>                                             0

and_vertebrates %>% filter(section=="CC") %>% group_by(species) %>% summarise(sum(is.na(weight_g))) #Nombre manquant

## # A tibble: 4 × 2
##   species                    `sum(is.na(weight_g))`
##   <chr>                                       <int>
## 1 Cascade torrent salamander                      4
## 2 Coastal giant salamander                     2378
## 3 Cutthroat trout                              4275
## 4 <NA>                                            2

Avec Tidyverse 2

and_vertebrates %>%
  filter(section=="CC") %>%
  group_by(species) %>% 
  summarise(mean=mean(weight_g,na.rm=T),
            sd=sd(weight_g,na.rm=T),
            min=min(weight_g,na.rm=T),
            q1=quantile(weight_g,probs=0.25,na.rm=T),
            mediane=quantile(weight_g,probs=0.5,na.rm=T),
            q3=quantile(weight_g,probs=0.75,na.rm=T),
            max=max(weight_g,na.rm=T),
            N=n(),
            n=sum(!is.na(weight_g)),
            missing=sum(is.na(weight_g)))

## # A tibble: 4 × 11
##   species     mean     sd    min    q1 mediane    q3     max     N     n missing
##   <chr>      <dbl>  <dbl>  <dbl> <dbl>   <dbl> <dbl>   <dbl> <int> <int>   <int>
## 1 Cascade …   1.32  0.472   0.8   0.94     1.3   1.6    1.95     9     5       4
## 2 Coastal …   9.81 11.5     0.29  3.4      6.7  11.5  103.    5406  3028    2378
## 3 Cutthroa…   9.38 10.5     0.09  1.3      6.6  13.4   77.8  11073  6798    4275
## 4 <NA>      NaN    NA     Inf    NA       NA    NA   -Inf        2     0       2

En une seule commande [base R]

tapply(and_vertebrates$weight_g[and_vertebrates$section=="CC"], and_vertebrates$species[and_vertebrates$section=="CC"], summary)

## $`Cascade torrent salamander`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.800   0.940   1.300   1.318   1.600   1.950       4 
## 
## $`Coastal giant salamander`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.290   3.400   6.700   9.811  11.500 102.990    2378 
## 
## $`Cutthroat trout`
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   0.090   1.300   6.600   9.381  13.400  77.780    4275

En une seule commande [tidyverse]

and_vertebrates %>% filter(section=="CC") %>% select(weight_g,species) %>% split(.$species) %>% map(summary)

## $`Cascade torrent salamander`
##     weight_g       species         
##  Min.   :0.800   Length:9          
##  1st Qu.:0.940   Class :character  
##  Median :1.300   Mode  :character  
##  Mean   :1.318                     
##  3rd Qu.:1.600                     
##  Max.   :1.950                     
##  NA's   :4                         
## 
## $`Coastal giant salamander`
##     weight_g         species         
##  Min.   :  0.290   Length:5406       
##  1st Qu.:  3.400   Class :character  
##  Median :  6.700   Mode  :character  
##  Mean   :  9.811                     
##  3rd Qu.: 11.500                     
##  Max.   :102.990                     
##  NA's   :2378                        
## 
## $`Cutthroat trout`
##     weight_g        species         
##  Min.   : 0.090   Length:11073      
##  1st Qu.: 1.300   Class :character  
##  Median : 6.600   Mode  :character  
##  Mean   : 9.381                     
##  3rd Qu.:13.400                     
##  Max.   :77.780                     
##  NA's   :4275

En beau

# [librairie gtsummary]
and_vertebrates %>% 
  filter(section=="CC") %>%
  tbl_summary(
    include = weight_g,
    by = species,
    type = all_continuous() ~ "continuous2",
    statistic = all_continuous() ~ c("{N_nonmiss}",
                                     "{N_miss}",
                                     "{min}",
                                     "{p25}",
                                     "{median}",
                                     "{p75}",
                                     "{max}",
                                     "{mean}",
                                     "{sd}"),
    missing = "no"
  ) %>%
  add_overall()

Characteristic	Overall, N = 16,488	Cascade torrent salamander, N = 9	Coastal giant salamander, N = 5,406	Cutthroat trout, N = 11,073
weight_g
N	9,831	5	3,028	6,798
N missing	6,657	4	2,378	4,275
Minimum	0	1	0	0
25%	2	1	3	1
Median	7	1	7	7
75%	13	2	12	13
Maximum	103	2	103	78
Mean	10	1	10	9
SD	11	0	11	10

En graphique

and_vertebrates %>% 
  filter(section=="CC") %>%
  drop_na(weight_g) %>% 
  ggplot() +
  aes(x=species, y=weight_g,color=species) +
  geom_boxplot() +
  labs(title="Plot of weight per species",x="species", y = "weight") +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

and_vertebrates %>% 
  filter(section=="CC") %>%
  drop_na(weight_g) %>% 
  ggplot() +
  aes(x=species, y=weight_g,color=species) +
  geom_boxplot() +
  labs(title="Plot of weight per species",x="species", y = "weight") +
  coord_flip()

Statistiques descriptives : moyenne, médiane,…

Librairies

Import du jeu de données and_vertebrates

Décrire une variable continue

Codes

Une commande par indicateur [base R]

Une commande par indicateur [Tidyverse]

En une seule commande

En joli

En graphique

Décrire une variable continue d’un sous-groupe de notre jeu de données

Codes

Une commande par indicateur [base R]

Une commande par indicateur [Tidyverse]

En une seule commande

En joli

En graphique

Décrire une variable continue en fonction d’une variable catégorielle

Codes

Une commande par indicateur [base R 1]

Une commande par indicateur [base R 2]

Avec Tidyverse 1

Avec Tidyverse 2

En une seule commande [base R]

En une seule commande [tidyverse]

En beau

En graphique

Décrire une variable continue en fonction d’une variable catégorielle d’un sous-groupe de notre jeu de données

Codes

Une commande par indicateur [base R 1]

Une commande par indicateur [base R 2]

Avec Tidyverse 1

Avec Tidyverse 2

En une seule commande [base R]

En une seule commande [tidyverse]

En beau

En graphique