Case study - gapminder

class: title-slide

# Case study - gapminder

## tidy models with `broom`

.center[<img src="https://raw.githubusercontent.com/tidymodels/broom/master/man/figures/logo.png" width="100px"/>]

### A. Ginolhac | rworkshop | 2021-09-10

<div>
<style type="text/css">.xaringan-extra-logo {
width: 110px;
height: 30px;
z-index: 0;
background-image: url(img/00/logo_uni_elixir.png);
background-size: contain;
background-repeat: no-repeat;
position: absolute;
bottom:1em;left:1em;
}
</style>
<script>(function () {
  let tries = 0
  function addLogo () {
    if (typeof slideshow === 'undefined') {
      tries += 1
      if (tries < 10) {
        setTimeout(addLogo, 100)
      }
    } else {
      document.querySelectorAll('.remark-slide-content:not(.title-slide):not(.inverse):not(.hide_logo)')
        .forEach(function (slide) {
          const logo = document.createElement('a')
          logo.classList = 'xaringan-extra-logo'
          logo.href = 'https://rworkshop.uni.lu/'
          slide.appendChild(logo)
        })
    }
  }
  document.addEventListener('DOMContentLoaded', addLogo)
})()</script>
</div>

---
# Learning objectives

.flex[
.w-70.bg-washed-green.b--green.ba.bw2.br3.shadow-5.ph3.mt3.ml1[
.large[.gbox[You will learn]]

.float-img[![](img/00/kt.png)]

.Large[

- Use `dplyr` / `purrr` for efficient data manipulation
- Tidying multiple linear models using `broom` 
- Managing related things together in **one** `tibble`
- Summarise findings in one `ggplot` using relevant aesthetics

]]

.w-30.bg-washed-yellow.b--yellow.ba.bw2.br3.shadow-5.ph3.mt3.ml1[
.large[.gbox[Guided practical]]

.center.Large[Interactive session]

]]

---

# Managing multiple models

#### Tutorial based on the great conference by [Hadley Wickham][1]

.center[![](img/gap_hadley_purrr.png)]

[1]:https://www.youtube.com/watch?v=rz3_FDVt9eg

---

# List-column cheatsheet, reminder

.flex[
.w-50.b--green.ba.bw1.br3.shadow-5.ph3.mt2.mr1[
.large[.ybox[`list` in a `data.frame/tibble`]]

.center[![](img/map_tibble_list_col_vectors.png)]
]
.w-50.b--green.ba.bw1.br3.shadow-5.ph3.mt2.mr1[
.large[.ybox[`list of tibbles` in a `data.frame/tibble`]]

.center[![](img/map_tibble_list_col_tibbles.png)]
]]

.center.footnote[[Functional Programming by S. Altman, B. Behrman, H. Wickham](https://dcl-prog.stanford.edu/list-columns.html) (CC4 Licence)]

---

# Keep all analyses together

### Steps: cols, per group: rows

#### Workflow from last time

.pull-left[

```r
palmerpenguins::penguins %>%
  nest_by(island, species) %>%
  mutate(model = list(lm(bill_depth_mm ~ bill_length_mm,
                         data = data)),
         summary = list(summary(model)),
         r_squared = pluck(summary, "r.squared"))
```

#### Remember

`nest_by()` is roughly the equivalent to:

```r
penguins %>%
  group_by(island, species) %>%
  summarise(data = list(cur_data())) %>%
* rowwise()
```

- `rowwise()` ensures that operations are performed on each row

]

.pull-right[

```
# A tibble: 5 × 6
# Rowwise:  island, species
  island    species                 data model  summary    r_squared
  <fct>     <fct>     <list<tibble[,6]>> <list> <list>         <dbl>
1 Biscoe    Adelie              [44 × 6] <lm>   <smmry.lm>    0.219 
2 Biscoe    Gentoo             [124 × 6] <lm>   <smmry.lm>    0.414 
3 Dream     Adelie              [56 × 6] <lm>   <smmry.lm>    0.258 
4 Dream     Chinstrap           [68 × 6] <lm>   <smmry.lm>    0.427 
5 Torgersen Adelie              [52 × 6] <lm>   <smmry.lm>    0.0620
```
]

---

# [Gapminder](http://www.gapminder.org/) is a fact tank

.pull-left[

## Dataset

.float-img[![](img/gap_jbryan.jpg)]

- From an independent Swedish foundation
- [R package](https://github.com/jennybc/gapminder) by [Jenny Bryan!](https://github.com/jennybc)
- Install from CRAN: `gapminder`
]

.pull-right[

## Hans Rosling

.float-img[![](img/gap_hrosling.jpg)]

- Fundamentally optimistic
- Great [talk](https://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen)
]

---
class: slide-practical

# Guided practical, explore gapminder

.flex[
.w-60.bg-washed-green.b--green.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.gbox[Questions <svg aria-hidden="true" role="img" viewBox="0 0 384 512" style="height:1em;width:0.75em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M202.021 0C122.202 0 70.503 32.703 29.914 91.026c-7.363 10.58-5.093 25.086 5.178 32.874l43.138 32.709c10.373 7.865 25.132 6.026 33.253-4.148 25.049-31.381 43.63-49.449 82.757-49.449 30.764 0 68.816 19.799 68.816 49.631 0 22.552-18.617 34.134-48.993 51.164-35.423 19.86-82.299 44.576-82.299 106.405V320c0 13.255 10.745 24 24 24h72.471c13.255 0 24-10.745 24-24v-5.773c0-42.86 125.268-44.645 125.268-160.627C377.504 66.256 286.902 0 202.021 0zM192 373.459c-38.196 0-69.271 31.075-69.271 69.271 0 38.195 31.075 69.27 69.271 69.27s69.271-31.075 69.271-69.271-31.075-69.27-69.271-69.27z"/></svg>]]

.large[
- Install the `gapminder` package
- Load `gapminder` and `tidyverse` packages
- Use the pipe `%>%` to pass `gapminder` to `ggplot()`
- **Plot** the `life expectency` (`lifeExp` in `y`) ~ `year` (`x`)
- Use `geom_line()`
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Warning!]]

.float-img[![](img/00/warning.png)]

.center.huge[mind the grouping! ]
]]

---

# Gapminder

.pull-left[

```r
# install.packages("gapminder")
library(gapminder)
gapminder %>%
  ggplot(aes(x = year, y = lifeExp,
*            group = country)) +
  geom_line()
```
]

.pull-right[
<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-4-1.png" width="504" />
]

---
class: slide-practical

# Keep related things together using list-column

.flex[
.w-60.bg-washed-green.b--green.ba.bw2.br3.shadow-5.ph3.mt3.ml6[
.large[.gbox[Questions <svg aria-hidden="true" role="img" viewBox="0 0 384 512" style="height:1em;width:0.75em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M202.021 0C122.202 0 70.503 32.703 29.914 91.026c-7.363 10.58-5.093 25.086 5.178 32.874l43.138 32.709c10.373 7.865 25.132 6.026 33.253-4.148 25.049-31.381 43.63-49.449 82.757-49.449 30.764 0 68.816 19.799 68.816 49.631 0 22.552-18.617 34.134-48.993 51.164-35.423 19.86-82.299 44.576-82.299 106.405V320c0 13.255 10.745 24 24 24h72.471c13.255 0 24-10.745 24-24v-5.773c0-42.86 125.268-44.645 125.268-160.627C377.504 66.256 286.902 0 202.021 0zM192 373.459c-38.196 0-69.271 31.075-69.271 69.271 0 38.195 31.075 69.27 69.271 69.27s69.271-31.075 69.271-69.271-31.075-69.27-69.271-69.27z"/></svg>]]

.large[
- Add a column using `mutate()` named `year1950` which is:  
`year` - 1950
- Nest with `nest_by()` the tibble by `country` and `continent`
- How many rows will you get?
.bold[Expectations help catching mistakes]
- Save the object as `by_country`
]
]
]

---

# Keep related things together

.pull-left[

### Nest _per_ country

```r
by_country <- gapminder %>%
  mutate(year1950 = year - 1950) %>%
  nest_by(continent, country)
by_country
```

### Helpers

- By default the list column is named `data`
- `year1950` will help to get meaningful intercepts
- Add `continent` to keep it along with `country`
- Note that `nest_by()` contains `rowwise()`

]
.pull-right[

```
# A tibble: 142 × 3
# Rowwise:  continent, country
   continent country                                data
   <fct>     <fct>                    <list<tibble[,5]>>
 1 Africa    Algeria                            [12 × 5]
 2 Africa    Angola                             [12 × 5]
 3 Africa    Benin                              [12 × 5]
 4 Africa    Botswana                           [12 × 5]
 5 Africa    Burkina Faso                       [12 × 5]
 6 Africa    Burundi                            [12 × 5]
 7 Africa    Cameroon                           [12 × 5]
 8 Africa    Central African Republic           [12 × 5]
 9 Africa    Chad                               [12 × 5]
10 Africa    Comoros                            [12 × 5]
# … with 132 more rows
```

]

---
class: slide-compact

# One country example: Germany

.pull-left[

### From original tibble

```r
gapminder %>%
  filter(country == "Germany") %>%
  select(-country, -continent)
```

```
# A tibble: 12 × 4
    year lifeExp      pop gdpPercap
   <int>   <dbl>    <int>     <dbl>
 1  1952    67.5 69145952     7144.
 2  1957    69.1 71019069    10188.
 3  1962    70.3 73739117    12902.
 4  1967    70.8 76368453    14746.
 5  1972    71   78717088    18016.
 6  1977    72.5 78160773    20513.
 7  1982    73.8 78335266    22032.
 8  1987    74.8 77718298    24639.
 9  1992    76.1 80597764    26505.
10  1997    77.3 82011073    27789.
11  2002    78.7 82350671    30036.
12  2007    79.4 82400996    32170.
```
]

.pull-right[

### Nested tibble

```r
by_country %>%
  filter(country == "Germany")
```

```
# A tibble: 1 × 3
# Rowwise:  continent, country
  continent country               data
  <fct>     <fct>   <list<tibble[,5]>>
1 Europe    Germany           [12 × 5]
```

```r
by_country %>%
  filter(country == "Germany") %>%
  unnest(data) # also summarise(data)
```

```
# A tibble: 12 × 7
# Groups:   continent, country [1]
   continent country  year lifeExp      pop gdpPercap year1950
   <fct>     <fct>   <int>   <dbl>    <int>     <dbl>    <dbl>
 1 Europe    Germany  1952    67.5 69145952     7144.        2
 2 Europe    Germany  1957    69.1 71019069    10188.        7
 3 Europe    Germany  1962    70.3 73739117    12902.       12
 4 Europe    Germany  1967    70.8 76368453    14746.       17
 5 Europe    Germany  1972    71   78717088    18016.       22
 6 Europe    Germany  1977    72.5 78160773    20513.       27
 7 Europe    Germany  1982    73.8 78335266    22032.       32
 8 Europe    Germany  1987    74.8 77718298    24639.       37
 9 Europe    Germany  1992    76.1 80597764    26505.       42
10 Europe    Germany  1997    77.3 82011073    27789.       47
11 Europe    Germany  2002    78.7 82350671    30036.       52
12 Europe    Germany  2007    79.4 82400996    32170.       57
```
]

---

# What happens in the `data.frame`, STAYS in the `data.frame`

.vembedr[
<div class="vembedr">
<div>
<iframe src="https://www.youtube.com/embed/4MfUCX_KpdE?start=943" width="533" height="300" frameborder="0" allowfullscreen=""></iframe>
</div>
</div>
]

---
class: slide-practical

# Las Vegas principle, add linear models

.large[
- Using `by_country`
- Add a new column `model` with linear regressions of `lifeExp` on `year1950`
- Save as `by_country_lm`
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Ask yourself]]

.float-img[![](img/00/warning.png)]

.center.Large[If you see **add column**, do you use `mutate` or `summarise`?]
]]

---

# Linear models

### Linear model _per_ country 
.pull-left[

```r
by_country_lm <- by_country %>%
  mutate(model = list(lm(lifeExp ~ year1950, data = data)))
by_country_lm
```
]

.pull-right[

```
# A tibble: 142 × 4
# Rowwise:  continent, country
   continent country                                data model 
   <fct>     <fct>                    <list<tibble[,5]>> <list>
 1 Africa    Algeria                            [12 × 5] <lm>  
 2 Africa    Angola                             [12 × 5] <lm>  
 3 Africa    Benin                              [12 × 5] <lm>  
 4 Africa    Botswana                           [12 × 5] <lm>  
 5 Africa    Burkina Faso                       [12 × 5] <lm>  
 6 Africa    Burundi                            [12 × 5] <lm>  
 7 Africa    Cameroon                           [12 × 5] <lm>  
 8 Africa    Central African Republic           [12 × 5] <lm>  
 9 Africa    Chad                               [12 × 5] <lm>  
10 Africa    Comoros                            [12 × 5] <lm>  
# … with 132 more rows
```
]

---
class: slide-practical

# Explore a list column

.flex[
.w-50.bg-washed-green.b--green.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.gbox[Questions <svg aria-hidden="true" role="img" viewBox="0 0 384 512" style="height:1em;width:0.75em;vertical-align:-0.125em;margin-left:auto;margin-right:auto;font-size:inherit;fill:currentColor;overflow:visible;position:relative;"><path d="M202.021 0C122.202 0 70.503 32.703 29.914 91.026c-7.363 10.58-5.093 25.086 5.178 32.874l43.138 32.709c10.373 7.865 25.132 6.026 33.253-4.148 25.049-31.381 43.63-49.449 82.757-49.449 30.764 0 68.816 19.799 68.816 49.631 0 22.552-18.617 34.134-48.993 51.164-35.423 19.86-82.299 44.576-82.299 106.405V320c0 13.255 10.745 24 24 24h72.471c13.255 0 24-10.745 24-24v-5.773c0-42.86 125.268-44.645 125.268-160.627C377.504 66.256 286.902 0 202.021 0zM192 373.459c-38.196 0-69.271 31.075-69.271 69.271 0 38.195 31.075 69.27 69.271 69.27s69.271-31.075 69.271-69.271-31.075-69.27-69.271-69.27z"/></svg>]]

.large[
- Count # rows per country using the `data` column
- Does any country have less data than others?
]
]

.w-50.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Reminder]]

.large[
- A **list column** is a list, you need to iterate through elements
- .bold[But] remember the `tibble` is still `rowwise` grouped (good).
- `distinct()` will help to find unique values
- .bold[But] remember the `tibble` is still `rowwise` grouped (.bold[bad]).
]]]

---

# Explore a list column

.flex[
.w-50.mr2[

```r
by_country_lm %>%
* mutate(n = nrow(data)) %>%
  select(continent, country, n)
```
]

.w-50.ml2[

```
# A tibble: 142 × 3
# Rowwise:  continent, country
   continent country                      n
   <fct>     <fct>                    <int>
 1 Africa    Algeria                     12
 2 Africa    Angola                      12
 3 Africa    Benin                       12
 4 Africa    Botswana                    12
 5 Africa    Burkina Faso                12
 6 Africa    Burundi                     12
 7 Africa    Cameroon                    12
 8 Africa    Central African Republic    12
 9 Africa    Chad                        12
10 Africa    Comoros                     12
# … with 132 more rows
```

]
]

.flex[
.w-50.mr2[

```r
by_country_lm %>%
  mutate(n = nrow(data)) %>%
  ungroup() %>% 
* distinct(n)
```
]

.w-50.ml2[

```
# A tibble: 1 × 1
      n
  <int>
1    12
```

]
]

---
class: slide-practical

# Explore a list column, plotting

.large[
- Plot `lifeExp` ~ `year1950` for **Bulgaria** by .bold[unnesting] `data`
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[reminder]]

.large[
- `filter()` for the desired country
- `unnest()` raw `data`
- Pipe to `ggplot()`]
]]

---

# Explore a list column, plotting

.pull-left[

```r
by_country_lm %>%
  filter(country == "Bulgaria") %>%
  unnest(data) %>%
  ggplot(aes(x = year1950, y = lifeExp)) +
  geom_line()
```
]

.pull-right[
<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-16-1.png" width="504" />

]

---
class: slide-practical

# Explore nested tibble

.large[
- Display the `summary` for the linear model of **Rwanda**
- How do you interpret the `\(r^2\)` for this particular model?
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[reminder]]

.large[
- `filter()` for the desired country
- Use `list()` to run `summary()` on the linear model
- To extract the named `"r.squared"`, use the `pluck(sumary, "r.squared")` `purrr` syntax]
]]

---

# Linear model for Rwanda

.pull-left[

```r
by_country_lm %>%
  filter(country == "Rwanda") %>%
  mutate(summary = list(summary(model)),
         r2 = pluck(summary, "r.squared")) %>% 
  select(country, r2)
```

```
Adding missing grouping variables: `continent`
```

```
# A tibble: 1 × 3
# Rowwise:  continent, country
  continent country     r2
  <fct>     <fct>    <dbl>
1 Africa    Rwanda  0.0172
```
]
.pull-right[
.large[
- `\(r^2\)` is close to 0, linearity sounds broken
- `broom` will cleanup linear model elements into tibbles
]
]

---

# Cleanup using `broom`

.center[![](img/gap_broom.png)]

---
class: slide-practical

# Tidying models

.large[
- Install `broom` from .bold[CRAN]
- Using `by_country_lm`, add 4 new columns:
    + `glance`, using the broom function on the `model` column
    + `tidy`, using the broom function on the `model` column
    + `augment`, using the broom function on the `model` column
    + `rsq` from the `glance` column
- Save as `models`
- Why extracting the `\(r^2\)` in the main tibble is useful?
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Reminder]]

.large[
- Use `list()` when dealing with a list column `rowwise` grouped
]]]

---

# Tidying models

.pull-left[

### Useful info

- Coefficients estimates:
    + **slope**
    + **intercept**
-  `\(r^2\)`
- Residuals

```r
library(broom)
models <- by_country_lm %>%
  mutate(glance  = list(glance(model)),
         tidy    = list(tidy(model)),
         augment = list(augment(model)),
         rsq     = pluck(glance, "r.squared"))
```

]

.pull-right[

### Extracting `\(r^2\)` in main tibble

.large[Why? no need to `unnest` for sorting / filtering.]

<div data-pagedtable="false">
  <script data-pagedtable-source type="application/json">
{"columns":[{"label":["continent"],"name":[1],"type":["fct"],"align":["left"]},{"label":["country"],"name":[2],"type":["fct"],"align":["left"]},{"label":["data"],"name":[3],"type":["list<tibble[,5]>"],"align":["right"]},{"label":["model"],"name":[4],"type":["list"],"align":["right"]},{"label":["glance"],"name":[5],"type":["list"],"align":["right"]},{"label":["tidy"],"name":[6],"type":["list"],"align":["right"]},{"label":["augment"],"name":[7],"type":["list"],"align":["right"]},{"label":["rsq"],"name":[8],"type":["dbl"],"align":["right"]}],"data":[{"1":"Africa","2":"Algeria","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98511721"},{"1":"Africa","2":"Angola","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.88781463"},{"1":"Africa","2":"Benin","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96660199"},{"1":"Africa","2":"Botswana","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.03402340"},{"1":"Africa","2":"Burkina Faso","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.91871050"},{"1":"Africa","2":"Burundi","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.76599597"},{"1":"Africa","2":"Cameroon","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.68017839"},{"1":"Africa","2":"Central African Republic","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.49324448"},{"1":"Africa","2":"Chad","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.87237550"},{"1":"Africa","2":"Comoros","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99685076"},{"1":"Africa","2":"Congo, Dem. Rep.","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.34820278"},{"1":"Africa","2":"Congo, Rep.","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.51966079"},{"1":"Africa","2":"Cote d'Ivoire","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.28337240"},{"1":"Africa","2":"Djibouti","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97437134"},{"1":"Africa","2":"Egypt","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99030424"},{"1":"Africa","2":"Equatorial Guinea","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99686864"},{"1":"Africa","2":"Eritrea","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95727334"},{"1":"Africa","2":"Ethiopia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96850263"},{"1":"Africa","2":"Gabon","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.81276621"},{"1":"Africa","2":"Gambia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98923562"},{"1":"Africa","2":"Ghana","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98409873"},{"1":"Africa","2":"Guinea","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97831518"},{"1":"Africa","2":"Guinea-Bissau","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98455829"},{"1":"Africa","2":"Kenya","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.44255729"},{"1":"Africa","2":"Lesotho","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.08485635"},{"1":"Africa","2":"Liberia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.51175640"},{"1":"Africa","2":"Libya","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98333149"},{"1":"Africa","2":"Madagascar","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99465364"},{"1":"Africa","2":"Malawi","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.83995446"},{"1":"Africa","2":"Mali","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99545140"},{"1":"Africa","2":"Mauritania","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99767430"},{"1":"Africa","2":"Mauritius","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.93478457"},{"1":"Africa","2":"Morocco","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99458312"},{"1":"Africa","2":"Mozambique","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.77427932"},{"1":"Africa","2":"Namibia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.43702163"},{"1":"Africa","2":"Niger","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.89768664"},{"1":"Africa","2":"Nigeria","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.87010508"},{"1":"Africa","2":"Reunion","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96607180"},{"1":"Africa","2":"Rwanda","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.01715964"},{"1":"Africa","2":"Sao Tome and Principe","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95525936"},{"1":"Africa","2":"Senegal","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99054417"},{"1":"Africa","2":"Sierra Leone","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96015054"},{"1":"Africa","2":"Somalia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.84442863"},{"1":"Africa","2":"South Africa","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.31246865"},{"1":"Africa","2":"Sudan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99214243"},{"1":"Africa","2":"Swaziland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.06821087"},{"1":"Africa","2":"Tanzania","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.76421876"},{"1":"Africa","2":"Togo","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.90580373"},{"1":"Africa","2":"Tunisia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98070422"},{"1":"Africa","2":"Uganda","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.34215382"},{"1":"Africa","2":"Zambia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.05983644"},{"1":"Africa","2":"Zimbabwe","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.05623196"},{"1":"Americas","2":"Argentina","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99556810"},{"1":"Americas","2":"Bolivia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98454156"},{"1":"Americas","2":"Brazil","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99804741"},{"1":"Americas","2":"Canada","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99638552"},{"1":"Americas","2":"Chile","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98279710"},{"1":"Americas","2":"Colombia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96787344"},{"1":"Americas","2":"Costa Rica","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96174767"},{"1":"Americas","2":"Cuba","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.92406716"},{"1":"Americas","2":"Dominican Republic","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97060781"},{"1":"Americas","2":"Ecuador","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99456626"},{"1":"Americas","2":"El Salvador","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95567201"},{"1":"Americas","2":"Guatemala","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99666377"},{"1":"Americas","2":"Haiti","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98761338"},{"1":"Americas","2":"Honduras","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97730026"},{"1":"Americas","2":"Jamaica","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.80565904"},{"1":"Americas","2":"Mexico","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98520444"},{"1":"Americas","2":"Nicaragua","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99677615"},{"1":"Americas","2":"Panama","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95119516"},{"1":"Americas","2":"Paraguay","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98298650"},{"1":"Americas","2":"Peru","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98847401"},{"1":"Americas","2":"Puerto Rico","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.90781912"},{"1":"Americas","2":"Trinidad and Tobago","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.79800744"},{"1":"Americas","2":"United States","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98592016"},{"1":"Americas","2":"Uruguay","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97683072"},{"1":"Americas","2":"Venezuela","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.94652607"},{"1":"Asia","2":"Afghanistan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.94771226"},{"1":"Asia","2":"Bahrain","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96673981"},{"1":"Asia","2":"Bangladesh","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98936087"},{"1":"Asia","2":"Cambodia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.63869222"},{"1":"Asia","2":"China","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.87127734"},{"1":"Asia","2":"Hong Kong, China","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97230183"},{"1":"Asia","2":"India","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96843652"},{"1":"Asia","2":"Indonesia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99711418"},{"1":"Asia","2":"Iran","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99501535"},{"1":"Asia","2":"Iraq","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.54578420"},{"1":"Asia","2":"Israel","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99478290"},{"1":"Asia","2":"Japan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95959563"},{"1":"Asia","2":"Jordan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96975008"},{"1":"Asia","2":"Korea, Dem. Rep.","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.70306306"},{"1":"Asia","2":"Korea, Rep.","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98765101"},{"1":"Asia","2":"Kuwait","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95235423"},{"1":"Asia","2":"Lebanon","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.94172582"},{"1":"Asia","2":"Malaysia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.94650639"},{"1":"Asia","2":"Mongolia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98731309"},{"1":"Asia","2":"Myanmar","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.87937750"},{"1":"Asia","2":"Nepal","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99154171"},{"1":"Asia","2":"Oman","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97479461"},{"1":"Asia","2":"Pakistan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99724965"},{"1":"Asia","2":"Philippines","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99142260"},{"1":"Asia","2":"Saudi Arabia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97208439"},{"1":"Asia","2":"Singapore","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98794751"},{"1":"Asia","2":"Sri Lanka","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.94771469"},{"1":"Asia","2":"Syria","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98416512"},{"1":"Asia","2":"Taiwan","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95707113"},{"1":"Asia","2":"Thailand","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96738440"},{"1":"Asia","2":"Vietnam","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98941189"},{"1":"Asia","2":"West Bank and Gaza","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97048087"},{"1":"Asia","2":"Yemen, Rep.","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98117240"},{"1":"Europe","2":"Albania","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.91057777"},{"1":"Europe","2":"Austria","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99213401"},{"1":"Europe","2":"Belgium","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99454056"},{"1":"Europe","2":"Bosnia and Herzegovina","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.89569829"},{"1":"Europe","2":"Bulgaria","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.54654217"},{"1":"Europe","2":"Croatia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.93243047"},{"1":"Europe","2":"Czech Republic","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.91668191"},{"1":"Europe","2":"Denmark","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97066797"},{"1":"Europe","2":"Finland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99383835"},{"1":"Europe","2":"France","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99762458"},{"1":"Europe","2":"Germany","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98950568"},{"1":"Europe","2":"Greece","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97196085"},{"1":"Europe","2":"Hungary","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.79501875"},{"1":"Europe","2":"Iceland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97032657"},{"1":"Europe","2":"Ireland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98414574"},{"1":"Europe","2":"Italy","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99336612"},{"1":"Europe","2":"Montenegro","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.80186521"},{"1":"Europe","2":"Netherlands","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98221566"},{"1":"Europe","2":"Norway","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96290057"},{"1":"Europe","2":"Poland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.83966315"},{"1":"Europe","2":"Portugal","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96903508"},{"1":"Europe","2":"Romania","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.80556666"},{"1":"Europe","2":"Serbia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.87880538"},{"1":"Europe","2":"Slovak Republic","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.79174822"},{"1":"Europe","2":"Slovenia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96604327"},{"1":"Europe","2":"Spain","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.96489456"},{"1":"Europe","2":"Sweden","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99548216"},{"1":"Europe","2":"Switzerland","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.99739086"},{"1":"Europe","2":"Turkey","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98533264"},{"1":"Europe","2":"United Kingdom","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.98443596"},{"1":"Oceania","2":"Australia","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.97964774"},{"1":"Oceania","2":"New Zealand","3":"<list<tibble[,5]>>","4":"<S3: lm>","5":"<tibble[,12]>","6":"<tibble[,5]>","7":"<tibble[,8]>","8":"0.95358464"}],"options":{"columns":{"min":{},"max":[10]},"rows":{"min":[10],"max":[10]},"pages":{}}}
  </script>
</div>

]

---
class: center, middle, inverse

# Exploratory plots

---
class: slide-practical

# Plotting `\(r^2\)` for countries

.large[
- Plot `country` ~ `rsq` 
- Color points per continent
- **Reorder** country levels by `\(r^2\)` (`rsq`): _snake plot_
- Which continent shows most of the low `\(r^2\)` values?
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Reminder]]

.large[
to reorder the discrete values of `country`:

- Use the `forcats` package
- `fct_reorder(country, rsq)` to reorder based on the `rsq` continuous variable
]]]

---

# Do linear models fit all countries?

### Snake plot

.pull-left[

```r
library(forcats)
models %>%
  ggplot(aes(x = rsq, 
             y = fct_reorder(country,
                             rsq))) +
  geom_point(aes(colour = continent), 
             alpha = 0.7, size = 2) +
  theme_classic(18) +
  theme(axis.text.y = element_blank(),
        axis.ticks.y = element_blank(),
        legend.position = c(0.25, 0.75)) +
   guides(color = guide_legend(
     override.aes = list(alpha = 1))) +
  labs(x = "r square",
       y = "Country") 
```
]
.pull-right[
<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-25-1.png" width="504" />

]

---
class: slide-practical

# Display the real data for countries with a low `\(r^2\)`

.large[
- Focus on non-linear trends
- Filter the 20 countries with the lowest `\(r^2\)` 
- `unnest` column `data`
- Plot `lifeExp` ~ `year` with lines
- Colour per continent
- Facet per country
- Same questions for the **top 20** `\(r^2\)`
]
]

.w-40.bg-washed-red.b--red.ba.bw2.br3.shadow-5.ph3.mt3.ml2[
.large[.rbox[Reminder]]

.large[
- You .bold[must] `ungroup()` as we work currently .bold[by row]
- `slice_min(col, n = 5)` returns the 5 minimal values of `col`
]]]

---

# Focus on non-linear trends

.pull-left[

```r
models %>%
  ungroup() %>% 
  slice_min(rsq, n = 20) %>%
  unnest(data) %>%
  ggplot(aes(x = year, y = lifeExp)) +
  geom_line(aes(colour = continent)) +
  facet_wrap(~ country) +
  theme(axis.text.x = element_text(angle = 45,
                                   hjust = 1),
        legend.position = "bottom")
```
]
.pull-right[
<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-27-1.png" width="504" />

]

---

# Focus on best linear trends

.pull-left[

```r
models %>%
  ungroup() %>% 
  slice_max(rsq, n = 20) %>%
  unnest(data) %>%
  ggplot(aes(x = year, y = lifeExp)) +
  geom_line(aes(colour = continent)) +
  facet_wrap(~ country) +
  theme(axis.text.x = element_text(angle = 45,
                                   hjust = 1),
        legend.position = "bottom")
```
]
.pull-right[
<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-28-1.png" width="504" />

]

---

# Interpreting the linear model

.pull-left[
### Regression

- What represents the **intercept**?
    + Using `year1950`?
    + Using `year`?
    + Justify Hadley choice
- What represents the **slope**?

<img src="lecture09_lm_gapminder_files/figure-html/unnamed-chunk-29-1.png" width="432" />
]

.pull-right[

- Coefficients with predictor `year - 1950`

```r
filter(models, country == "Germany") %>%
  unnest(tidy) %>%
  select(continent, country, estimate)
```

```
# A tibble: 2 × 3
# Groups:   continent, country [1]
  continent country estimate
  <fct>     <fct>      <dbl>
1 Europe    Germany   67.1  
2 Europe    Germany    0.214
```
- Compare with model and original years

```r
gapminder %>%
  filter(country == "Germany") %>%
  lm(lifeExp ~ year, data = .) %>% 
  tidy() %>%
  select(term, estimate)
```

```
# A tibble: 2 × 2
  term        estimate
  <chr>          <dbl>
1 (Intercept) -350.   
2 year           0.214
```

]

---
class: slide-practical

# Summarise on one plot

### by Hadley Wickham

- Unnest coefficients (`tidy` column)
    + Mind to keep the `continent`, `country` and `rsq` columns
- Put **intercept** and **slope** in their own columns
    + In **wide** format, only one value can be used. 
    + Discard unused columns.
- Plot `slope ~ intercept` (watch out the `(Intercept)` name which needs to be called between backticks '.large[`]')
- Colour per continent
- Size per `\(r^2\)` (use for `scale_size_area()` for lisibility)
- Add tendency with `geom_smooth(method = "loess")`
]
]

---

---
class: nvs1

count: false
 
### Full pipeline
.panel1-final_plot-auto[

```r
*gapminder
```
]
 
.panel2-final_plot-auto[

```
# A tibble: 1,704 × 6
   country     continent  year lifeExp      pop gdpPercap
   <fct>       <fct>     <int>   <dbl>    <int>     <dbl>
 1 Afghanistan Asia       1952    28.8  8425333      779.
 2 Afghanistan Asia       1957    30.3  9240934      821.
 3 Afghanistan Asia       1962    32.0 10267083      853.
 4 Afghanistan Asia       1967    34.0 11537966      836.
 5 Afghanistan Asia       1972    36.1 13079460      740.
 6 Afghanistan Asia       1977    38.4 14880372      786.
 7 Afghanistan Asia       1982    39.9 12881816      978.
 8 Afghanistan Asia       1987    40.8 13867957      852.
 9 Afghanistan Asia       1992    41.7 16317921      649.
10 Afghanistan Asia       1997    41.8 22227415      635.
# … with 1,694 more rows
```
]