Why?
Before building any model you have to understand your data. Exploratory data analysis and good graphics are how data scientists find problems, patterns and hypotheses worth testing.
What?
Exploratory data analysis in R: the data analysis process, the principles of analytic graphics, exploratory graphs, R plotting systems, data manipulation, clustering, and techniques for working with large datasets. A real air pollution dataset is used throughout.
Curriculum:
Data Analysis Overview
What data analysis is, types of data, types of analysis, tools and the data analysis process.
Exploratory Data Analysis in R
Formulating a question, reading in data, checking the packaging, str(), looking at the top and bottom, checking your n's, validating with external data and challenging your solution.
Principles of Analytic Graphics
Show comparisons, causality and mechanism, multivariate data, integrate evidence, and document the evidence.
Exploratory Graphs
Characteristics of exploratory graphs, one- and two-dimensional summaries, and the air pollution use case.
Plotting Systems in R
The base, lattice and ggplot2 plotting systems and when to use each.
Data Manipulation
Selecting, filtering, arranging, grouping and summarising data in R.
Data Clustering
Hierarchical clustering and k-means, plus dimension reduction ideas.
Working with Large Datasets
Reading, storing and processing large datasets efficiently in R.
Notes
Hands-on course in R. You need a working R environment and basic R syntax. The course project asks you to reproduce plots and analyse a dataset, graded on the plots, the analysis and your code.