Syllabus Intro to R

ECON 890: Intro to R

Class time: Mo/Tu/Th 9:30am - 12:00pm

This class composes a six work course Introduction to R. The main objective is to give you tools that are helpful in research. We will favor breath over depth. Ideally, the schedule for the course will be the following:

• Lecture 1: Basics in R: R packages, basic operations, file management, objects in R, classes, tests and conversion.
• Lecture 2: R Markdown: Where we will do mosts homeworks (read this post: http://www.stat.cmu.edu/~cshalizi/rmarkdown/ ; we will go through it in the second class).
• Lecture 3: Functions and Linear Regressions: statistical functions, descriptive statistics, writing functions, conditional statements and loops.
• Lecture 4: Good coding practices (I) and Data work: creating, merging, exporting and importing datasets; missing values; working with large datasets; strings and dates, tidyr, tibble and dplyr.
• Lecture 5: Graphs in R: base graphics package, ggplot2 package; leaftlet.
• Lecture 6: More on regressions (plot, splines, time series, quantile regression, tests)
• Lecture 7: Causal inference in R
• Lecture 8: Text analysis: http://tidytextmining.com/
• Lecture 9: Networks in R and DAGS
• Lecture 10: Numerical optimization: different methods.
• Lecture 11: Good coding practices (II) and R performance: coding tips, R packages for efficient coding, paralleling R, debugging code.

If you have some suggestions for topics, please, contact me.

Assessment

The grades will be determined by problem sets (50%) and a project (50%). The main aim of problem sets is to advance your knowledge of R. You should complete the problems individually, but you can discuss the homeworks with your peers. In fact, I recommend you to talk to your classmates. However, copying and pasting code of your classmates is prohibited. After the homework submission, we will spend some time discussing the solutions in class.

Projects

For your project split in groups of 2 to 4 people. You will need to replicate a paper of your choosing. You should send me an email with the proposed paper, where you explained if the data is obtainable. I will then approve your paper, if it seems likely you will be able to replicate it (you should convince me it is doabl; perhaps you should consider choosing an AER paper, as they have datasets available for the published papers). If you have no way of finding the data, you could simulate it. You should replicate the main results of the paper (of course, if you do not have the real data, you should use the same methods but you are not expected to deliver the same numerical results). Submit a write up of results and the code¹. The RMarkdown file (master file) should compile the html or latex file, and explain how result is arising, along with the underlying code and the corresponding relevant output. The folder should also contain any additional files that are needed to obtain the results in your write-up. Deadline for submission is July 27, at 11:59pm.

Submission guidelines

The idea of going through RMarkdown was also to simplify delivery of assignments and readability of code. Thus, please construct each assignment in html or pdf format, with both the code and the relevant output. One file should suffice. Use the following naming convention: Lastname_A’X’.(html/pdf). ‘X’ stands for the current number of days where a homework had to be delivered. For the first file, X=1, for the second, X=2, and so on. Please upload it to your Drop Box folder on Sakai. You should upload the final project there as well.

---

1. Write up should be done in R Markdown and presented in html or latex format.↩