- Instructor: Randy Lai
You should use Campuswire or Canvas to contact me. DO NOT send email to me as I tend to ignore emails (too much spams). - Meeting time: 12:10 - 1:30 PM, TR
- TA: Hoseung Song, Cong Xu
- Office hour:
- Hoseung: 2:00 - 4:00pm PM Thursday
- Cong: 1:00 - 3:00pm PM Friday
- Randy: 10:00 AM - 12:00 PM Friday
| Date | Note | HTML |
|---|---|---|
| 01-05 | Introduction | introduction |
| 01-07 | functional programming | purrr |
| 01-19 | debug | debug |
| profile | profile | |
| 01-25 | Rcpp | Rcpp |
| 02-02 | R-pkgs | |
| 02-04 | parallel | parallel |
| 02-10 | bootstrap | bootstrap |
| blb | blb | |
| 02-18 | random forest | rf |
| 02-25 | python | python |
| 03-02 | keras | keras |
- Canvas for grades
- GitHub for lecture notes and assignments. Please always refer to the documents found on GitHub.
- Campuswire for discussions.
Please use your ucd email to register! If you have used piazza
before, it is similar but has a better interface. Please use
Campuswire to ask any questions related to assignments and course
materials. I and the TA will not answer any emails related to the
course materials.
Use Join code: 0149
Learn how to ask a question. Asking a question is an art, stackoverflow.com has some good tips. You could also use the reprex package to make reproducible examples.
| Topic |
|---|
| Introduction |
| Functional programming |
| Debugging and profiling |
| Writing C and C++ extensions |
| Writing an R package |
| Parallel & Distributed Computing |
| Bootstrap and BLB |
| Interoperate with Python and Julia |
| Deep learning in R |
| Category | Grade Percentage |
|---|---|
| Assignments | 65% |
| Final Project | 25% |
| Participation | 10% |
- There will be around 6 assignments and they are assigned via GitHub classroom.
- Assignments must be turned in by the due date. No late assignments are accepted.
- Participation will be based on your reputation point in Campuswire.
- 1% each week if the reputation point for the week is above 20.
- the top scorers for the quarter will earn extra bonuses.
- J. Bryan, the STAT 545 TAs, J. Hester, Happy Git and GitHub for the useR (https://happygitwithr.com/)
- J. Bryan, Data wrangling, exploration, and analysis with R (https://stat545.com/)
- G. Grolemund and H. Wickham, R for Data Science (https://r4ds.had.co.nz/)
- H. Wickham, Advanced R (https://adv-r.hadley.nz/)
- And a lot of google
- Strong in R programming
- R 4.0.3 (check your R version)
- RStudio 1.3.1093 (check your RStudio Version)
- R Markdown (read this https://rmarkdown.rstudio.com/lesson-1.html)
- Knowledge about git and GitHub: read ‘Happy Git and GitHub for the useR’ (It is absoluately important to read the ebook if you have no experiences with git/GitHub)
- Some knowledge of dplyr will be helpful
- Minimal amount of Python
Assuming that you have git installed,
- Open RStudio -> New Project -> Version Control -> Git -> paste the URL: https://github.com/ucdavis-sta141c-2021-winter/sta141c-lectures.git
- Choose a directory to create the project
- You could make any changes to the repo as you wish.
- To fetch updates
- go to the
gitpane in RStudio - click the “Commit” button and
- check the files changed by you
- type a short message about the changes and hit “Commit”
- After committing the message, hit the “Pull” button (PS: there is a sub button “Pull with rebase”, only use it if you truly understand what it is)
- Done if you see no errors
- If there were lines which are updated by both me and you, you would see a merge conflict.
- To resolve the conflict, locate the files with conflicts (U flag in the git pane).
- Open the files and edit the conflicts, usually a conflict looks like
<<<<<<< HEAD - RStudio 1.2.5011 (check your RStudio Version) ======= - RStudio 1.2.5033 (check your RStudio Version) >>>>>>> 85858c9a6ebba9057ca8db7c269bd0a2f7a3910a- check all the files with conflicts and commit them again with a new message.
- go to the
Link your github account at https://signin-apd27wnqlq-uw.a.run.app/sta141c/
Check regularly the course github organization https://github.com/ucdavis-sta141c-2021-winter for any newly posted assignments.
Feedback will be given in forms of GitHub issues or pull requests.
Regrade requests must be made within one week of the return of the assignment. One of the most common reasons is not having the knitted html files uploaded, 30% of the grade of that assignment will be deducted if it happens. To make a request, send me a Canvas message with the following information:
- Which assignment
- URL to the repo of your assignment
- The reason of the request
(Adapted from Nick Ulle and Clark Fitzgerald )
Point values and weights may differ among assignments. This is to indicate what the most important aspects are, so that you spend your time on those that matter most. Check the homework submission page on Canvas to see what the point values are for each assignment.
The grading criteria are correctness, code quality, and communication. The following describes what an excellent homework solution should look like:
The report does the following:
- solves all the questions contained in the prompt
- makes conclusions that are supported by evidence in the data
- discusses efficiency and limitations of the computation
- cites any sources used
The attached code runs without modification.
The code is idiomatic and efficient. Different steps of the data processing are logically organized into scripts and small, reusable functions. Variable names are descriptive. The style is consistent and easy to read.
Plots include titles, axis labels, and legends or special annotations where appropriate. Tables include only columns of interest, are clearly explained in the body of the report, and not too large. Writing is clear, correct English.
The report points out anomalies or notable aspects of the data discovered over the course of the analysis. It discusses assumptions in the overall approach and examines how credible they are. It mentions ideas for extending or improving the analysis or the computation.