02-plots.Rmd

# EXAMINING A DISTRIBUTION


**Chapter Links**

* [Chapter 2 Slide Show](http://tysonbarrett.com/EDUC-6600/Slides/u01_Ch2_DataViz.html#1)

**Assingment Links**

* [Unit 1 Assignment - Write Up Skeleton](https://usu.box.com/s/bn0m44cp5zrkev3ppjdilfsvic4bubvu)

* [Unit 1 Assignment - Rmd Skeleton](https://usu.box.com/s/zjjwexn827ziifxlcoe6emr2brfifk0x)

* [Inho's Dataset - Excel format](https://usu.box.com/s/hyky7eb24l6vvzj2xboedhcx1xolrpw1)



```{r global_options, include=FALSE}
# set global chunk options...  
#  this changes the defaults so you don't have to repeat yourself
knitr::opts_chunk$set(comment     = NA,
                      cache       = TRUE,
                      echo        = TRUE, 
                      warning     = FALSE, 
                      message     = FALSE)
```

## Example: Ihno's Experiment {-}

This is the same dataset used in the previous chapter and in most of the textbook's section C questions.

### Required Packages {-}

```{r load_libraries}
library(tidyverse)    # Loads several very helpful 'tidy' packages
library(readxl)        # Read in Excel datasets
library(furniture)    # Nice tables (by our own Tyson Barrett)
library(psych)        # Lots of nice tid-bits
```


### Data Import and Wrangling {-}

The previous chapter build the following code to read in the dataset and prepare it for analysis.

```{r, eval=FALSE}
data_clean <- read_excel("Ihno_dataset.xls") %>% 
  dplyr::rename_all(tolower) %>% 
  dplyr::mutate(genderF = factor(gender, 
                                 levels = c(1, 2),
                                 labels = c("Female", 
                                            "Male"))) %>% 
  dplyr::mutate(majorF = factor(major, 
                                levels = c(1, 2, 3, 4,5),
                                labels = c("Psychology",
                                           "Premed",
                                           "Biology",
                                           "Sociology",
                                           "Economics"))) %>% 
  dplyr::mutate(reasonF = factor(reason,
                                 levels = c(1, 2, 3),
                                 labels = c("Program requirement",
                                            "Personal interest",
                                            "Advisor recommendation"))) %>% 
  dplyr::mutate(exp_condF = factor(exp_cond,
                                   levels = c(1, 2, 3, 4),
                                   labels = c("Easy",
                                              "Moderate",
                                              "Difficult",
                                              "Impossible"))) %>% 
  dplyr::mutate(coffeeF = factor(coffee,
                                 levels = c(0, 1),
                                 labels = c("Not a regular coffee drinker",
                                            "Regularly drinks coffee")))  %>% 
  dplyr::mutate(hr_base_bps  = hr_base / 60) %>% 
  dplyr::mutate(anx_sum      = rowsums(anx_base, anx_pre, anx_post)) %>% 
  dplyr::mutate(hr_mean      = rowmeans(hr_base + hr_pre + hr_post)) %>% 
  dplyr::mutate(statDiff     = statquiz - exp_sqz)
```

```{r, include=FALSE}
data_clean <- read_excel("data/Ihno_dataset.xls") %>% 
  dplyr::rename_all(tolower) %>% 
  dplyr::mutate(genderF = factor(gender, 
                                 levels = c(1, 2),
                                 labels = c("Female", 
                                            "Male"))) %>% 
  dplyr::mutate(majorF = factor(major, 
                                levels = c(1, 2, 3, 4,5),
                                labels = c("Psychology",
                                           "Premed",
                                           "Biology",
                                           "Sociology",
                                           "Economics"))) %>% 
  dplyr::mutate(reasonF = factor(reason,
                                 levels = c(1, 2, 3),
                                 labels = c("Program requirement",
                                            "Personal interest",
                                            "Advisor recommendation"))) %>% 
  dplyr::mutate(exp_condF = factor(exp_cond,
                                   levels = c(1, 2, 3, 4),
                                   labels = c("Easy",
                                              "Moderate",
                                              "Difficult",
                                              "Impossible"))) %>% 
  dplyr::mutate(coffeeF = factor(coffee,
                                 levels = c(0, 1),
                                 labels = c("Not a regular coffee drinker",
                                            "Regularly drinks coffee")))  %>% 
  dplyr::mutate(hr_base_bps  = hr_base / 60) %>% 
  dplyr::mutate(anx_sum      = rowsums(anx_base, anx_pre, anx_post)) %>% 
  dplyr::mutate(hr_mean      = rowmeans(hr_base + hr_pre + hr_post)) %>% 
  dplyr::mutate(statDiff     = statquiz - exp_sqz)
```


```{r}
tibble::glimpse(data_clean)
```

---------------------------------------

## Frequency Tables

These tables are best for showing the breakdown of a sample across the levels of a **single CATEGORICAL** variable *(factor)*.  They help pick out the mode(s) and identify unusual or impossible values [@R-furniture].

```{r}
data_clean %>%                 
  furniture::tableF(majorF)
```

Phobia is one variable that is in between being categorical and continuous.

```{r}
data_clean %>% 
  furniture::tableF(phobia)
```

If a variable has many possible values (i.e. it is more continuous than categorical), you can add an option to tell how many values `n = #` you want displayed in the table, cutting out all the middle values.

```{r}
data_clean %>% 
  furniture::tableF(hr_post, n = 10)
```


---------------------------------------

## Bar Charts

These plots are best for showing the breakdown of a sample across the levels of a **single CATEGORICAL** variable.  They help pick out the mode(s) and identify unusual or impossible values [@R-ggplot2].

> There must be SPACE between the bars!

```{r}
data_clean %>% 
  ggplot(aes(majorF)) +
  geom_bar()
```

Here is an example of a two-level variable

```{r}
data_clean %>% 
  ggplot(aes(coffeeF)) +
  geom_bar()
```

Here is an example of an 11-level variable

```{r}
data_clean %>% 
  ggplot(aes(phobia)) +
  geom_bar()
```

--------------------------------

## Histograms

These plots are best for showing the distribution of a **single CONTINUOUS** variable.  They help visually determine the shape, center *[mean, median, mode(s)]*, spread *[stadard deviation, range]*, and identify extreme or impossible values [@R-ggplot2].

```{r}
data_clean %>% 
  ggplot(aes(phobia)) +
  geom_histogram()
```

> There must NOT be SPACE between the bars!

Notice how the bars do not touch.  This is because the default includes too many bars, many of which are not included in the dataset.  

There are TWO ways specify something other than the default:

(@) Change the NUMBER of bins: `bins = #`
(@) Change the WIDTH of the bins: `binwidth = #`

> If you try to do BOTH, only the first option will be used and the second will be ignored.

### Change the Number of Bins

```{r}
data_clean %>% 
  ggplot(aes(phobia)) +
  geom_histogram(bins = 8)
```


### Change the Bin Width

```{r}
data_clean %>% 
  ggplot(aes(phobia)) +
  geom_histogram(binwidth = 5)
```



### Make Seperate Panels -by- a Factor

To make separate plots based on another categorical variable, a **FACTOR**, we need to add a layer to the plot.

> Reminder: Steps before the `ggplot()` are combined with pipes `%>%`, whereas layers of the plot are combined with the addition symbol `+`.


```{r}
data_clean %>% 
  ggplot(aes(mathquiz)) +
  geom_histogram(binwidth = 4) +
  facet_grid(. ~ coffeeF)
```

```{r}
data_clean %>% 
  ggplot(aes(mathquiz)) +
  geom_histogram(binwidth = 4) +
  facet_grid(coffeeF ~ .)
```


------------------------------

## Percentiles

The `quantile(probs = c(#, #, ..., #))` function in the base download of R may be used to request the deciles, quartiles, or other percentiles for a `vector` for numbers [@R-base].

To use this function, we have to first **pull** out one variable of interest from our dataset *(data.frame)* and make it into a vector.  This is done with a `dplyr::pull(varname)` step [@R-dplyr].


### Deciles 

Deciles break the variable's values into 10% chunks.

```{r}
data_clean %>% 
  dplyr::pull(statquiz) %>% 
  quantile(probs = c(0, .10, .20, .30, .40, .50, .60, .70, .80, .90, 1))
```


#### With Missing Values {-}

If the variable have any missing values, an error message with be outputted instead of what you expect.

```{r, eval = FALSE}
data_clean %>% 
  dplyr::pull(mathquiz) %>% 
  quantile(probs = c(0, .10, .20, .30, .40, .50, .60, .70, .80, .90, 1))
```

`Error in quantile.default(., probs = c(0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, : missing values and NaN's not allowed if 'na.rm' is FALSE`

#### Option to Ignore Missing Values {-}

To avoid getting this message and ignore the missing values, use the `na.rm = TRUE` option.

```{r}
data_clean %>% 
  dplyr::pull(mathquiz) %>% 
  quantile(probs = c(0, .10, .20, .30, .40, .50, .60, .70, .80, .90, 1),
           na.rm =TRUE)
```




### Quartiles 

Quartiles break variable's values into 4 chunks.  If you include 0 *(minimum)* and 1 *(maximum)* you will get the **Five Number Summary**, which is:

* min, the minimum
* Q1, the 25th percentile
* Q2, the 50th percentile or the Median
* Q3, the 75th percentile
* max, the maximum

These values are used to create boxplots in the next chapter.

```{r}
data_clean %>% 
  dplyr::pull(statquiz) %>% 
  quantile(probs = c(0, .25, .50, .75, 1))
```




### Other Percentiles

You may also include any other percentile between 0 and 1.

```{r}
data_clean %>% 
  dplyr::pull(statquiz) %>% 
  quantile(probs = c(.01, .05, .173, .90))
```