It is great that you are all ready to learn Advanced Data Analysis in R which is starting tomorrow (every Saturday, starting from October 17, 2020) via Zoom at 4 pm GMT or 5 pm BST/UK time or 6 pm CET/European time. We will only use 1 hour for each class meeting and the video recordings of each lesson on Zoom will be uploaded on YouTube. Please remember to subscribe, share and click the YouTube notification bell for future YouTube uploads on Advanced Data Analysis. So don't worry if you miss any section or whatever we do at each meeting. The main goal is to develop the interest of mathematical programming among students from Africa. However, anyone who interested is welcome to join and learn. This class will be a good introduction to data analysis with advanced statistical software like R.

Below are the details of tasks to be accomplished for our first & second class:

A brief introduction into mathematical programming and the advantages of R to other programming languages (Python, SAS, etc.) and other statistical software like SPSS, STATA, EVIEWS, MINITAB, GRETL, etc.
Expatiate on how R works, setting up working directories and importing data into R (Note that you can import a data saved as .csv, .txt, .xlsx, SPSS data, STATA data, SAS data, etc). Writing codes as a script, saving scripts and importing previously saved scripts for use another time. You be using R studio for everything; but after the first class, I will rather be using Jupyter notebook to run R whiles you use the R studio. The many importance of Jupyter notebook as an IDE to run R, Python and other programming languages will be explained in our second meeting.
How to use in-built functions, install & load packages for some specific statistical analysis, and create your own function(s) for any complicated task depending on what you want. We will create our own function to compute for instance correlation between two quantitative variables based on a formula, and compare the results with the in-built correlation function in R. This will help you to know that any complicated formula can be programmed and evaluated in R.
We shall analyse our first real data where we will create our own novel function to compute summary statistics of the data (see the Resource section to download the CSV data named as School_data.csv). We will learn how to calculate BMI as an additional variable if we already have height and weight as variables in the data. This is just to help you to know how to create additional variables in your data based on existing variables. You will learn how to convert a quantitative variable (BMI) into a qualitative variable (BMI status with categories: underweight, normal weight, overweight and obese). We will learn how to assign names to categorical variables coded with dummy numbers (eg. 0,1,2 etc). The novel summary statistics function we will be creating has never been created before. First, it must be automated such that for any data you use, it should be able to do the following. i) determine the type of any/all variable(s) in the data (whether categorical/qualitative or numerical/quantitative), ii) If a variable is numerical (like age, BMI, height,etc.), it should compute & return the mean, median, standard deviation, standard error, skewness and kurtosis; iii) else if a variable is categorical (like gender, marital status, etc.) it should return its percentages for all categories/levels of the variable rounded to exactly 1 decimal place. You will be given an assignment/task to add more complexity to this novel summary statistics function we shall create in class to be able to include descriptive plots as well.
Lastly, we shall do some data visualizations in R (bar plots, histograms, pie chart, line graphs and qqplots).

NB: Create a folder on your desktop & name it DataAnalysis_results_R and download & put the CSV file/data named School_data.csv into it (this data is at the Resource section).

Zoom link for the meetings every Saturday
https://cardiff.zoom.us/j/3811424245?pwd=UWhLdzRKU0xsNW1PSXhmU0dCVmNjQT09

Meeting ID: 381 142 4245
Password: 474761

#pin