Visualizing Linear Models

# Visualizing Linear Models
### Dr. D’Agostino McGowan

---

<div class="my-footer">
<span>
Dr. Lucy D'Agostino McGowan
</span>
</div>

---

## Visualizing linear models

* Forest plots of the coefficients and confidence intervals

![](23-visualizing-linear-models_files/figure-html/unnamed-chunk-2-1.png)

---

## Building up to this plot!

First, fit the model and tidy it

```r
library(broom)
library(ggplot2)
mod <- lm(mpg ~ cyl + am + vs + gear, data = mtcars)
mod_tidy <- tidy(mod, conf.int = TRUE)
mod_tidy <- mod_tidy[-1, ] # drop the intercept
```

---

## Building up to this plot!

First, fit the model and tidy it

```r
library(broom)
library(ggplot2)
*mod <- lm(mpg ~ cyl + am + vs + gear, data = mtcars)
mod_tidy <- tidy(mod, conf.int = TRUE)
mod_tidy <- mod_tidy[-1, ] # drop the intercept
```

---

## Building up to this plot!

First, fit the model and tidy it

```r
library(broom)
library(ggplot2)
mod <- lm(mpg ~ cyl + am + vs + gear, data = mtcars)
*mod_tidy <- tidy(mod, conf.int = TRUE)
mod_tidy <- mod_tidy[-1, ] # drop the intercept
```

---

## Building up to this plot!

First, fit the model and tidy it

```r
library(broom)
library(ggplot2)
mod <- lm(mpg ~ cyl + am + vs + gear, data = mtcars)
mod_tidy <- tidy(mod, conf.int = TRUE)
*mod_tidy <- mod_tidy[-1, ] # drop the intercept
```

---

## Create a forest plot

```r
ggplot(mod_tidy,
       aes(x = term, y = estimate, 
           ymin = conf.low,
           ymax = conf.high)) +
* geom_pointrange()
```
]
]

--
.question[
What could be better?
]

---

## Flip the x- and y-axis

```r
ggplot(mod_tidy,
       aes(x = term, y = estimate, 
           ymin = conf.low,
           ymax = conf.high)) +
  geom_pointrange() +
* coord_flip()
```
]
]

---

## Axis labels

```r
ggplot(mod_tidy,
       aes(x = term, y = estimate, 
           ymin = conf.low,
           ymax = conf.high)) +
  geom_pointrange() +
  coord_flip() +
* labs(x = "",
*      y = "Coefficient (95% CI)")
```
]
]

---

## Term labels

```r
mod_tidy$labels <- c(
* "No. cylinders",
* "Manual transmission",
* "Straight engine",
* "No. forward gears")
ggplot(mod_tidy,
*      aes(x = labels, y = estimate,
           ymin = conf.low,
           ymax = conf.high)) +
  geom_pointrange() +
  coord_flip() +
  labs(x = "", 
       y = "Coefficient (95% CI)") 
```
]
]

Better ways to do this (especially if you have lots of variables!)

* `merge()`
* `left_join()`

---

## Add a reference line at the null

```r
ggplot(mod_tidy,
       aes(x = labels, y = estimate, 
           ymin = conf.low,
           ymax = conf.high)) +
  geom_pointrange() +
* geom_hline(yintercept = 0, lty = 2) +
  coord_flip() +
  labs(x = "", 
       y = "Coefficient (95% CI)") 
```
]
]

---

## Update the theme

```r
ggplot(mod_tidy,
       aes(x = labels, y = estimate, 
           ymin = conf.low,
           ymax = conf.high)) +
  geom_pointrange() +
  geom_hline(yintercept = 0, lty = 2) +
  coord_flip() +
  labs(x = "", 
       y = "Coefficient (95% CI)") +
* theme_minimal()
```
]
]

* Check out the available themes here: https://ggplot2.tidyverse.org/reference/ggtheme.html

---

## Higlight points of interest

```r
ggplot(mod_tidy,
       aes(x = labels, y = estimate, 
           ymin = conf.low,
           ymax = conf.high,
*          color =
*            ifelse(p.value < 0.05,
*                   "p < 0.05",
*                   "p > 0.05"))) +
  geom_pointrange() +
  geom_hline(yintercept = 0, lty = 2) +
  coord_flip() +
* scale_color_manual("",
*                    values = c(
*                      "blue",
*                      "black")) +
  labs(x = "", 
       y = "Coefficient (95% CI)") +
  theme_minimal() 
```
]
]

---

## Annotate an important point

```r
ggplot(mod_tidy,
       aes(x = labels, y = estimate, 
           ymin = conf.low,
           ymax = conf.high,
           color = 
             ifelse(p.value < 0.05, 
                    "p < 0.05", 
                    "p > 0.05"))) + 
  geom_pointrange() +
  geom_hline(yintercept = 0, lty = 2) +
  coord_flip() +
  scale_color_manual("", 
                     values = c( 
                       "blue", 
                       "black")) + 
  labs(x = "", 
       y = "Coefficient (95% CI)") +
* annotate("text", x = 2.25, y = 5.9,
*          label = "Having manual transmission \n(versus automatic) \nyeilds an expected change in \nmiles per gallon of 4.32 \n(95% CI: 0.1, 8.5)",
*          size = 2) +
  theme_minimal() 
```
]
]

---

## Annotate an important point

```r
ggplot(mod_tidy,
       aes(x = labels, y = estimate, 
           ymin = conf.low,
           ymax = conf.high,
           color = 
             ifelse(p.value < 0.05, 
                    "p < 0.05", 
                    "p > 0.05"))) + 
  geom_pointrange() +
  geom_hline(yintercept = 0, lty = 2) +
  coord_flip() +
  scale_color_manual("", 
                     values = c( 
                       "blue", 
                       "black")) + 
  labs(x = "", 
       y = "Coefficient (95% CI)") +
  annotate("text", x = 2.25, y = 5.9,
           label = "Having manual transmission \n(versus automatic) \nyeilds an expected change in \nmiles per gallon of 4.32 \n(95% CI: 0.1, 8.5) holding all other \nvariables constant", 
           size = 2) + 
* annotate("curve", x = 1.5, y = 4.75,
*          xend = 1.2, yend = 4.32,
*          arrow = arrow(length = unit(0.03, "npc"))) +
  theme_minimal() 
```
]
]

---

## And there you have it!

![](23-visualizing-linear-models_files/figure-html/unnamed-chunk-25-1.png)

---

## <svg style="height:0.8em;top:.04em;position:relative;" viewBox="0 0 640 512"><path d="M624 416H381.54c-.74 19.81-14.71 32-32.74 32H288c-18.69 0-33.02-17.47-32.77-32H16c-8.8 0-16 7.2-16 16v16c0 35.2 28.8 64 64 64h512c35.2 0 64-28.8 64-64v-16c0-8.8-7.2-16-16-16zM576 48c0-26.4-21.6-48-48-48H112C85.6 0 64 21.6 64 48v336h512V48zm-64 272H128V64h384v256z"/></svg>`Application Exercise`

* Take the model you chose in the previous application exercise and create a figure like the one discussed today. Include a caption that explains what the figure demonstrates.