Let’s Get Started

class: center, middle, inverse, title-slide

# Let’s Get Started
### Stephen Balogun
### October 11, 2021

---

## What this course will teach you?

.left-column[

- Setting up your workspace

- Navigating RStudio

- Overview of data science

- A quick revision of basic biostatistics

- Concept of tidy data

]

.right-column[

- R functions

- R packages

- Assigning values to an object

- R object (data) basic building blocks (vectors)

- Arithmetic operators

- Putting it together using simulated data - demo
]

---

## Who is this class for?

.left-column[

- New R users

- No prior computer programming knowledge

- No assumed prior knowledge of statistics

- No prior understanding of technical jargons

- People who work in the development sector or interested in research
]

.right-column[

<img src="./imgs/assumptions.png" width="2083" />
]

---
class: inverse, middle, center
background-image: url(https://upload.wikimedia.org/wikipedia/commons/3/39/Naruto_Shiki_Fujin.svg)
background-size: contain

# Let's get to know you!

---

## Setting up your workspace

.pull-left[

- Modify default settings

- From the __*Tools*__ tab, go to __*Global Options...*__ (the last listed item)

- Confirm that the places highlighted are unchecked

<img src="./imgs/workspace_1.png" width="3432" />
]

.pull-right[

<img src="./imgs/workspace2.png" width="2157" />
]

---

## Setting up your workspace - Project

.pull-left[

- Use R projects

- Makes your work easy to share with others

- Identified by a "__*.Rproj*__" file in the folder

- Create R project

- Click on the __*File*__ tab, then

- Click on __*New project*__ (should be the second option from the listed items)

- Create from "New" or "Existing" folders

]

.pull-right[

]

---

## Navigating RStudio

.left-column[

- 3 - 4 window panes

- Use the Editor pane for your codes

- Send codes from the __*editor*__ to the __*console*__ either using the __*run*__ tab (it has a drop-down) or use "Ctrl + Enter"

]

.right-column[

<img src="./imgs/rstudio pane.png" width="1702" />
]

---

## Data Science - Overview

.left-column[
- A combination of

- computer programming
    
    - Scientific reasoning 
    
    - Data comprehension
    
- Varying mix of the above, depending on the specialization

]

.pull-right[
<img src="./imgs/datascience component2.png" width="560" />
]

---

## Data Science - Workflow

.left-column[

- Getting your data (Excel, csv, etc) into R

- Formatting your data appropriately

- Identifying patterns (exploration)

- Plotting graphs

- Performing statistics

- Passing the information

]

.pull-right[
<img src="./imgs/datascience workflow.png" width="1336" />
]

---

## Biostats overview - tabular data

.left-column[

- Rows are horizontal

- Columns are vertical

- Should be __*tidy*__

- Each variable must have its own column
    
    - Each observation must have its own row
    
    - Each value must have its own cell
]

.right-column[
<img src="./imgs/tidy data.png" width="1600" />

> "Tidy datasets are alike, but every messy dataset is messy in its own way"
  >
  > - Hadley Wickham
]

---

## Biostats overview - tidy data

.pull-left[

What makes this data *"untidy"*?
<img src="./imgs/untidy data.png" width="728" />
]

.pull-right[
What makes this data "*tidy"*?

<img src="./imgs/tidy.png" width="766" />
]

---

## Biostats overview - variables

- Numeric variables

- Continuous (can be expressed as fractions) e.g. weight
    
    - Discrete (always whole numbers) e.g. heart rate

- Categorical variables (generally coded)

- ordinal (a graduation pattern observable) e.g. level of education
    
    - nominal (no ranking) e.g. gender/sex
    
- Text variables e.g. Name

- __*Type of analysis depends on the variable type*__

- __*Variables can be converted from one form to another*__ e.g. age (a numeric variable), can be binned into age-groups (a categorical variable)

---

## Biostats overview - charts

- The type of chart to use depend on the variable type

- Common graphs include:

- bar charts (one categorical variable)
    
    - histograms (continuous variable)
    
    - scatter-plots (two continuous variables)
    
    - line plots (one continuous and one categorical/discrete vs one continuous)
    
    - box-plots (Used for visualizing summary statistics of continuous variable +/- categorical variable)
    
    - pie charts (avoid as much as possible)

---

## Biostatistics overview - statistics

- descriptive statistics

- measures of central tendencies (mean, median mode)
    
    - measures of dispersion (sd, IQR, range, variance, etc)
    
- inferential statistics

- involves test of hypothesis
    
[**More on introduction to biosatistics here!!!**](https://stbalo2002.shinyapps.io/Intro_to_biostatistics/)

---
class: inverse, middle, center
background-image: url(https://upload.wikimedia.org/wikipedia/commons/3/39/Naruto_Shiki_Fujin.svg)
background-size: contain

# Back to R!

---

## R functions

- Essentially the "heart" of R

- R can perform simple arithmetics like `1 + 1`

- Most functions follow this format:

function_name(object, 1st argument, 2nd argument, etc)

- Example, assuming you have a population of 10 and you want to sample 2 persons (without replacement)

sample(x = 1:10, size = 2, replace = FALSE)

- You can have functions with no arguments e.g. `Sys.Date()`

- You can also have functions without name (so called anonymous functions)

---

## R packages

.pull-left[

- All R functions are contained in *packages*

- *R packages* are a collection of *functions* that perform similar tasks

- R comes with pre-installed packages

- Packages extend the functions of base R

- Additional packages can be installed depending on your needs

- *CRAN* and *GitHub* are two common package repositories

- You can also write your own packages!!!

]

.pull-right[
- to install a package from CRAN, use:

`install.packages("pkg_name")`

- better still, use the `p_install()` function from the {pacman} package:

p_install("pkg_name")
    
- Load packages (except for base R packages)

- To load a package, use:

`library("pkg_name")` or
  
  `p_load("pkg_name")` # the `p_load()` function is from the {pacman} package so yu will have to load {pacman} first
  
]

---

## Assigning values

.pull-left[

- R can let you "save" a value for re-use

- To save a value, decide on the name you want to give it

- Use syntactic names

- A combination of letters, digits, ".", and "_"

- Don't start with a number
  
  - When you start with a ".", it cannot be followed by a number 
  
  - no spaces allowed
  
  - avoid *reserved* words
  
]

.pull-right[

- Use meaningful names (for your sake and others)

- Generally two common naming conventions (snakecase & camelcase):

- Here is a snakecase example: this_is_a_snake_case
  
  - Here is a camelcase example: thisIsACamelCase
  
- I recommend the snakecase

- Caution!! Existing values of a similar name will be overwritten so names have to be unique

]
  
---

## Vectors: data building blocks

.pull-left[

- How R recognise variable/data type

- Atomic vectors (4 common)

1. integers (equivalent of discrete variables) e.g. `c(20L, 4L, 100L)`
  
  2. double (equivalent of continuous variables) e.g. `c(20, 4, 100)`
  
      are approximations (is 2 equals to squaroot of 2 exponential 2?)
  
  3. logical/boolean
  
  4. characters e.g. `c("Stephen", "12", "TRUE")`
  
  5. complex
  
  6. raw
]

.pull-right[

#### Hierarchy of vectors

*try*: c(20, TRUE, "Stephen")
]

---

## Vectors: lists & augmented vectors

.pull-left[

How about dates & categorical variables?

- 4 Augmented vectors

1. factors (from integers)
  
  2. date (from double) 
  
    R format for date is the ISO8601 (yyyy-mm-dd)
  
  3. datetime (from double)
  
  4. duration i.e. time difference (from double)

]

.pull-right[

What if I want to have a mix of data types?

- Lists

e.g. `list(29, TRUE, "Stephen")`

- Dataframes

- data with rows and columns
    
    - variables are in the columns, observations in the rows (tidy data)
    
    - columns must be of equal length
    
    - each column can have only one type of vector

]

---

## Working with vectors

.pull-left[

What type of vector am I dealing with?

- use `typeof()` e.g.

`typeof(10)`
    
    `typeof("Stephen")`
    
    `typeof(TRUE)`
    
    `typeof("FALSE")` [practice]

]

.pull-right[
    
Can I check to know if my vector is of a pacticular type?

- use `is.atomic_type` e.g.

`is.double(10)`
    
    `is.logical(TRUE)`

Can I force my vectors to change from one type to another? yes

- use `as.atomic_type` e.g.

`as.integer("10")`
  
    `as.logical("FALSE")`

]

---

## Arithmetic operators

.pull-left[

- addition sign (+)

- substraction sign (-)

- division sign (/)

- multiplication sign (*)

- Exponential sign (^)

- integer division (%/%)

- modulo division (%%) - remainder

- greater than (>) e.g. 5 > 3

- less than (<) e.g. 5 < 3

]

.pull-right[

- greater than or equal to (>=) e.g. `5 >= 10/2`

- less than or equal to (<=) e.g. `5 <= 10/2`

- equal to (==) e.g. `1/49 * 49 == 1`

- not equal to (!=) e.g. `4 != 2*2`

- not/negate (!) e.g. `!TRUE`

- or (|) e.g. `4 > 6 | 3`

- and (&) e.g. `4 > 6 & 3`

- contains (%in%) e.g. `c("Stephen", "Taiye", "Kehinde", "Tade", "Tola") %in% c("Taiye", "Kehinde")`
]

---

## Putting it all together

.left-column[

- Simulated dataset (public health, lab or research)

- View [**demo data here**](https://raw.githubusercontent.com/stephenbalogun/intro_to_R_demo/master/data/demo_data.csv)

- A csv file containing 10,000 rows and 10 columns

- Real life data is not always this clean!!!

]

.right-column[

#### Data form
<img src="./imgs/form.png" width="1635" />

]

---

## Exploratory questions

1. How many persons from each state?

2. Plot a graph of the gender distribution

3. Plot a graph to demonstrate the relationship between weight and height

4. How many persons have systolic hypertension?

5. For each gender in each state, what are the average age, average height, and average weight?

---

## What have we learnt today?

1. Setting up the R work environment

2. Tidy data

3. Basic building blocks of R objects

4. Naming conventions in R

5. Performing basic arithmetics

6. Sourcing and using relevant R functions
---
class: inverse, middle, center
background-image: url(https://upload.wikimedia.org/wikipedia/commons/3/39/Naruto_Shiki_Fujin.svg)
background-size: contain

# Questions