5 Eating habits of fleas

Last modified on 06. June 2026 at 20:42:16

“A quote.” — Dan Meyer

5.1 General background

What is a data grid?

5.2 Theoretical background

5.3 R packages used

R Code [show / hide]

pacman::p_load(tidyverse, conflicted)

5.4 Data

Small data and grid

5.4.1 Fleas with feeding habits

C. felis felis jump was $19.9 \pm 9.1cm$ with a range from $2$ to $48cm$

C. canis jump was longer $30.4 \pm 9.1cm$ with a range from $3$ to $50cm$

The jump length of P. irritans has to be derived from the work of Cadiergues et al. (2000)² because no mean or standard deviation for the human flea was available.

Miarinjara (2025) et al.³ demonstrate the impact of sheep’s blood on the jumping ability of human fleas.

R Code [show / hide]

jump_flea_grid <- expand_grid(feeding = c("sugar water", "human blood")) |> 
  mutate(mean = c(19.9, 30.4), 
         sd = c(9.1))

R Code [show / hide]

jump_flea_tbl <- jump_flea_grid |> 
  rowwise() |> 
  mutate(jumplength = lst(rnorm(7, mean, sd))) |> 
  unnest(cols = jumplength) |> 
  mutate(feeding = as_factor(feeding))

R Code [show / hide]

jump_flea_tbl |> 
  group_by(feeding) |> 
  summarise(mean(jumplength),
            sd(jumplength)) |> 
  mutate_if(is.numeric, round, 2)

# A tibble: 2 × 3
  feeding     `mean(jumplength)` `sd(jumplength)`
  <fct>                    <dbl>            <dbl>
1 sugar water               21.2             7.3 
2 human blood               33.0             6.57

R Code [show / hide]

jump_animals_grid <- expand_grid(feeding = c("sugar water", "tomato ketchup", "fake blood", "human blood")) |> 
  mutate(mean = c(19.9, 35.2, 15.2, 30.4), 
         sd = c(9.1, 10.3, 4.6, 9.1))

R Code [show / hide]

jump_animals_tbl <- jump_animals_grid |> 
  rowwise() |> 
  mutate(jumplength = lst(rnorm(7, mean, sd))) |> 
  unnest(cols = jumplength) |> 
  mutate(feeding = as_factor(feeding))

R Code [show / hide]

jump_animals_tbl |> 
  group_by(feeding) |> 
  summarise(mean(jumplength),
            sd(jumplength)) |> 
  mutate_if(is.numeric, round, 2)

# A tibble: 4 × 3
  feeding        `mean(jumplength)` `sd(jumplength)`
  <fct>                       <dbl>            <dbl>
1 sugar water                  21.2             9.36
2 tomato ketchup               34.6            10.2 
3 fake blood                   13.4             2.51
4 human blood                  34.0            17.4

5.4.2 Fleas at stages of development

Why so complex?

⁴

R Code [show / hide]

jump_habitat_grid <- expand_grid(feeding = 1:3, stage = 1:2) |> 
  mutate(mean_feeding = c(19.9, 19.9, 30.4, 30.4, 15.2, 15.2),
         mean_stage = c(5, 0, 5, 0, 5, -5),
         mean = mean_feeding + mean_stage,
         sd = c(9.1, 9.1, 9.1, 9.1, 4.6, 4.6))
jump_habitat_grid

# A tibble: 6 × 6
  feeding stage mean_feeding mean_stage  mean    sd
    <int> <int>        <dbl>      <dbl> <dbl> <dbl>
1       1     1         19.9          5  24.9   9.1
2       1     2         19.9          0  19.9   9.1
3       2     1         30.4          5  35.4   9.1
4       2     2         30.4          0  30.4   9.1
5       3     1         15.2          5  20.2   4.6
6       3     2         15.2         -5  10.2   4.6

R Code [show / hide]

jump_habitat_raw_tbl <- jump_habitat_grid |> 
  rowwise() |> 
  mutate(jumplength = lst(rnorm(7, mean, sd))) |> 
  unnest(cols = jumplength)

R Code [show / hide]

jump_habitat_tbl <- jump_habitat_raw_tbl |> 
  select(feeding, stage, jumplength) |> 
  mutate(feeding = factor(feeding, labels = c("sugar water", "tomato ketchup", "human blood")),
         stage = factor(stage, labels = c("juvenile", "adult"))) |> 
  mutate_if(is.numeric, round, 2)

R Code [show / hide]

jump_habitat_tbl |> 
  group_by(feeding, stage) |> 
  summarise(mean(jumplength),
            sd(jumplength)) |> 
  mutate_if(is.numeric, round, 2)

# A tibble: 6 × 4
# Groups:   feeding [3]
  feeding        stage    `mean(jumplength)` `sd(jumplength)`
  <fct>          <fct>                 <dbl>            <dbl>
1 sugar water    juvenile               25.1            13.8 
2 sugar water    adult                  14.9            11.4 
3 tomato ketchup juvenile               29.9             6.32
4 tomato ketchup adult                  27.3             9.26
5 human blood    juvenile               22.0             3.65
6 human blood    adult                  10.1             4.6

5.4.3 Measurements on animal fleas

Jump length in [cm] called jump_length
Number of hairs on each leg called count_leg_left and count_leg_right
Ratings of each flea, as listed in the catalog of the Fédération Internationale de la Beauté des Puces (FIBP) called rating on a Likert scale from 1 to 5, with 5 being the strongest expression.
The infection status with the flea cold is called infected, with a value of 0/1 for no/yes.

R Code [show / hide]

tibble(jumplength = rnorm(7, 5, 1),
       counthairleg_left = rpois(7, 4),
       counthairleg_right = rpois(7, 4),
       counthairleg = (counthairleg_left + counthairleg_right)/2,
       rating = sample(1:5, 7, replace = TRUE, prob = c(0.1, 0.2, 0.4, 0.2, 0.1)),
       infectd = rbinom(7, prob = 0.5, size = 1))

# A tibble: 7 × 6
  jumplength counthairleg_left counthairleg_right counthairleg rating infectd
       <dbl>             <int>              <int>        <dbl>  <int>   <int>
1       5.02                 3                  5          4        2       0
2       4.73                 6                  4          5        3       0
3       5.91                 4                  5          4.5      3       0
4       3.38                 6                  3          4.5      2       1
5       3.10                 3                  4          3.5      5       0
6       7.02                 4                  3          3.5      1       1
7       3.95                 4                  3          3.5      1       0

5.5 Data availability

The data is available as txt-Files under https://github.com/jkruppa/biodatascience.

5.6 Alternatives

Further tutorials and R packages on XXX

5.6.1 Why linear is a bad idea?

R Code [show / hide]

jump_flea_tbl <- tibble(feeding = rep(0:1, each = 7),
                        jump_length = 19.9 + 10.5 * feeding + rnorm(14, 0, 9.1)) |> 
  mutate(feeding = factor(feeding, labels = c("cat", "dog")))

5.6.2 Why is a `tibble()` bad idea?

R Code [show / hide]

jump_animals_tbl <- tibble(cat = rnorm(n = 7, mean = 19.9, sd = 9.1),
                           fox = rnorm(n = 7, mean = 35.2, sd = 10.3),
                           rat = rnorm(n = 7, mean = 15.2, sd = 4.6),
                           dog = rnorm(n = 7, mean = 30.4, sd = 9.1)) |>
  pivot_longer(cols = cat:dog, values_to = "jump_length", names_to = "feeding") |> 
  mutate(feeding = as_factor(feeding))

5.7 Glossary

term: what does it mean.

5.8 The meaning of “Models of Reality” in this chapter.

itemize with max. 5-6 words

5.9 Summary

References

[1]

Wickham H, Çetinkaya-Rundel M, Grolemund G. R for Data Science: Import, Tidy, Transform, Visualize, and Model Data. " O’Reilly Media, Inc."; 2023.

[2]

Cadiergues MC, Joubert C, Franc M. A comparison of jump performances of the dog flea, ctenocephalides canis (curtis, 1826) and the cat flea, ctenocephalides felis felis (bouché, 1835). Veterinary parasitology. 2000;92(3):239-241.

[3]

Miarinjara A, Raveloson AO, Rajaonarimanana M, Ayala D, Girod R, Gillespie TR. Establishing a laboratory colony of the human flea, pulex irritans: Methods for collecting, rearing, and feeding. Parasites & Vectors. 2025;18(1):363.

[4]

Wickham H, Henry L. Purrr: Functional Programming Tools.; 2025. https://purrr.tidyverse.org/

```{r echo = FALSE, warning = FALSE, message=FALSE} source("init.R") source("images/part_1/part_1_data.R") ``` # Eating habits of fleas {#sec-data-examples} *Last modified on `r format(fs::file_info("chapter-12-factorial-data.qmd")$modification_time, '%d. %B %Y at %H:%M:%S')`* > *"A quote." --- Dan Meyer* ## General background What is a data grid? ## Theoretical background ```{r} #| message: false #| echo: false #| warning: false #| fig-align: center #| fig-height: 4 #| fig-width: 7 #| fig-cap: "foo" #| label: fig-factorial-data-overview p_factorial_data_overview ``` ```{r} #| message: false #| echo: false #| warning: false #| fig-align: center #| fig-height: 4 #| fig-width: 7 #| fig-cap: "foo" #| label: fig-covariate-data-overview p_covariate_data_overview ``` ## R packages used ```{r} #| message: false #| warning: false pacman::p_load(tidyverse, conflicted) ``` @wickham2023r ## Data Small data and grid ### Fleas with feeding habits @cadiergues2000comparison ```{r} #| echo: false #| message: false #| warning: false #| label: fig-chap-01-flea #| fig-align: center #| fig-height: 2 #| fig-width: 7 #| fig-cap: "foo **(A)** foo **(B)** foo" flea_jump_exp_p ``` *C. felis felis* jump was $19.9 \pm 9.1cm$ with a range from $2$ to $48cm$ *C. canis* jump was longer $30.4 \pm 9.1cm$ with a range from $3$ to $50cm$ The jump length of *P. irritans* has to be derived from the work of Cadiergues et al. (2000) @cadiergues2000comparison because no mean or standard deviation for the human flea was available. Miarinjara (2025) et al. @miarinjara2025establishing demonstrate the impact of sheep's blood on the jumping ability of human fleas. ```{r} jump_flea_grid <- expand_grid(feeding = c("sugar water", "human blood")) |> mutate(mean = c(19.9, 30.4), sd = c(9.1)) ``` ```{r} jump_flea_tbl <- jump_flea_grid |> rowwise() |> mutate(jumplength = lst(rnorm(7, mean, sd))) |> unnest(cols = jumplength) |> mutate(feeding = as_factor(feeding)) ``` ```{r} jump_flea_tbl |> group_by(feeding) |> summarise(mean(jumplength), sd(jumplength)) |> mutate_if(is.numeric, round, 2) ``` ```{r} #| echo: false #| message: false #| warning: false #| label: fig-chap-11-fac-01 #| fig-align: center #| fig-height: 3.5 #| fig-width: 7 #| fig-cap: "foo **(A)** foo **(B)** foo" jump_flea_tbl |> ggplot(aes(x = feeding, y = jumplength, fill = feeding)) + theme_book() + geom_point(position = position_nudge(-0.075), show.legend = FALSE, shape = 21) + stat_summary(fun.data=mean_sdl, fun.args = list(mult = 1), geom="pointrange", shape = 23, size = 1, show.legend = FALSE) + stat_summary(fun = mean, geom = "label", aes(label = round(..y..,2)), size = 4, position = position_nudge(x = 0.05), alpha = 0.5, show.legend = FALSE, hjust = "left") + labs(x = "Flea feeding", y = "Jump length in [cm]", fill = "feeding") + scale_fill_viridis(discrete = TRUE, option = "plasma") + coord_cartesian(ylim = c(0, NA)) ``` ```{r} jump_animals_grid <- expand_grid(feeding = c("sugar water", "tomato ketchup", "fake blood", "human blood")) |> mutate(mean = c(19.9, 35.2, 15.2, 30.4), sd = c(9.1, 10.3, 4.6, 9.1)) ``` ```{r} jump_animals_tbl <- jump_animals_grid |> rowwise() |> mutate(jumplength = lst(rnorm(7, mean, sd))) |> unnest(cols = jumplength) |> mutate(feeding = as_factor(feeding)) ``` ```{r} jump_animals_tbl |> group_by(feeding) |> summarise(mean(jumplength), sd(jumplength)) |> mutate_if(is.numeric, round, 2) ``` ```{r} #| echo: false #| message: false #| warning: false #| label: fig-chap-11-fac-02 #| fig-align: center #| fig-height: 3.5 #| fig-width: 7 #| fig-cap: "foo **(A)** foo **(B)** foo" jump_animals_tbl |> ggplot(aes(x = feeding, y = jumplength, fill = feeding)) + theme_book() + geom_point(position = position_nudge(-0.075), show.legend = FALSE, shape = 21) + stat_summary(fun.data=mean_sdl, fun.args = list(mult = 1), geom="pointrange", shape = 23, size = 1, show.legend = FALSE) + stat_summary(fun = mean, geom = "label", aes(label = round(..y..,2)), size = 4, position = position_nudge(x = 0.075), alpha = 0.5, show.legend = FALSE, hjust = "left") + labs(x = "Flea feeding", y = "Jump length in [cm]", fill = "feeding") + scale_fill_viridis(discrete = TRUE, option = "plasma") + coord_cartesian(ylim = c(0, NA)) ``` ### Fleas at stages of development Why so complex? @rpkg_purrr ```{r} jump_habitat_grid <- expand_grid(feeding = 1:3, stage = 1:2) |> mutate(mean_feeding = c(19.9, 19.9, 30.4, 30.4, 15.2, 15.2), mean_stage = c(5, 0, 5, 0, 5, -5), mean = mean_feeding + mean_stage, sd = c(9.1, 9.1, 9.1, 9.1, 4.6, 4.6)) jump_habitat_grid ``` ```{r} jump_habitat_raw_tbl <- jump_habitat_grid |> rowwise() |> mutate(jumplength = lst(rnorm(7, mean, sd))) |> unnest(cols = jumplength) ``` ```{r} jump_habitat_tbl <- jump_habitat_raw_tbl |> select(feeding, stage, jumplength) |> mutate(feeding = factor(feeding, labels = c("sugar water", "tomato ketchup", "human blood")), stage = factor(stage, labels = c("juvenile", "adult"))) |> mutate_if(is.numeric, round, 2) ``` ```{r} #| warning: false #| message: false jump_habitat_tbl |> group_by(feeding, stage) |> summarise(mean(jumplength), sd(jumplength)) |> mutate_if(is.numeric, round, 2) ``` ```{r} #| echo: false #| message: false #| warning: false #| label: fig-chap-11-fac-03 #| fig-align: center #| fig-height: 3.5 #| fig-width: 7 #| fig-cap: "foo **(A)** foo **(B)** foo" jump_habitat_tbl |> ggplot(aes(x = feeding, y = jumplength, fill = stage)) + theme_book() + geom_point(position = position_dodge(0.5), show.legend = FALSE, shape = 21) + stat_summary(fun.data=mean_sdl, fun.args = list(mult = 1), geom="pointrange", shape = 23, size = 1, show.legend = TRUE, position = position_dodge(0.9)) + stat_summary(fun = mean, geom = "label", aes(label = round(..y..,2)), size = 4, position = position_dodge(1.8), alpha = 0.5, show.legend = FALSE) + labs(x = "Flea feeding", y = "Jump length in [cm]", fill = "Development stage") + scale_fill_viridis(discrete = TRUE, option = "plasma") + coord_cartesian(ylim = c(0, NA)) ``` ### Measurements on animal fleas - Jump length in [cm] called `jump_length` - Number of hairs on each leg called `count_leg_left` and `count_leg_right` - Ratings of each flea, as listed in the catalog of the Fédération Internationale de la Beauté des Puces (FIBP) called `rating` on a Likert scale from 1 to 5, with 5 being the strongest expression. - The infection status with the flea cold is called `infected`, with a value of 0/1 for no/yes. ```{r} tibble(jumplength = rnorm(7, 5, 1), counthairleg_left = rpois(7, 4), counthairleg_right = rpois(7, 4), counthairleg = (counthairleg_left + counthairleg_right)/2, rating = sample(1:5, 7, replace = TRUE, prob = c(0.1, 0.2, 0.4, 0.2, 0.1)), infectd = rbinom(7, prob = 0.5, size = 1)) ``` ## Data availability The data is available as txt-Files under <https://github.com/jkruppa/biodatascience>. ```{r} #| echo: false jump_flea_tbl |> arrange(feeding) |> mutate_if(is.numeric, round, 2) |> write_delim("data/jump_fleas.txt") jump_animals_tbl |> arrange(feeding) |> mutate_if(is.numeric, round, 2) |> write_delim("data/jump_animals.txt") jump_habitat_tbl |> arrange(feeding, stage) |> mutate_if(is.numeric, round, 2) |> write_delim("data/jump_habitat.txt") ``` ## Alternatives Further tutorials and R packages on XXX ### Why linear is a bad idea? ```{r} jump_flea_tbl <- tibble(feeding = rep(0:1, each = 7), jump_length = 19.9 + 10.5 * feeding + rnorm(14, 0, 9.1)) |> mutate(feeding = factor(feeding, labels = c("cat", "dog"))) ``` ### Why is a `tibble()` bad idea? ```{r} jump_animals_tbl <- tibble(cat = rnorm(n = 7, mean = 19.9, sd = 9.1), fox = rnorm(n = 7, mean = 35.2, sd = 10.3), rat = rnorm(n = 7, mean = 15.2, sd = 4.6), dog = rnorm(n = 7, mean = 30.4, sd = 9.1)) |> pivot_longer(cols = cat:dog, values_to = "jump_length", names_to = "feeding") |> mutate(feeding = as_factor(feeding)) ``` ## Glossary term : what does it mean. ## The meaning of "Models of Reality" in this chapter. - itemize with max. 5-6 words ## Summary ## References {.unnumbered}

5.1 General background

5.2 Theoretical background

5.3 R packages used

5.4 Data

5.4.1 Fleas with feeding habits

5.4.2 Fleas at stages of development

5.4.3 Measurements on animal fleas

5.5 Data availability

5.6 Alternatives

5.6.1 Why linear is a bad idea?

5.6.2 Why is a tibble() bad idea?

5.7 Glossary

5.8 The meaning of “Models of Reality” in this chapter.

5.9 Summary

References

5.6.2 Why is a `tibble()` bad idea?