luosanj / stats507_f21 Goto Github PK

This is a github repository which I will used to share code and notebooks with students in Stats 507.

The primary page for this course, including static versions of slides, is available at jbhender.github.io/Stats507/F21.

The folder nb collects notebooks from the course, including lecture slides. Lecture slides are included both in the markdown .md format I authored them in and an associated .ipynb format viewable without jupytext enabled.

For the "tables" notebook, I've also included a version of the document as a python script. This was created using the command: jupytext --to py:light tables.ipynb. Note that these files have extra metadata, e.g # + [markdown], relative to what yours might. This metadata is here to enable me to show these as slides.

Files related to problem sets, including example solutions can be found in the ps folder.

Within the problem set (ps) folder, is a sub-folder "exemplars" containing student solutions that stood out as particulary good. Thank you to the students who agreed to have their solutions posted.

Demonstrations can be found in the demo folder. Currently these are:

• jupy_demo.py
• numpy_comp.py
• isolet_demo.py..

Within demo the shell directory contains full length demos and short examples illustrating syntax for useful shell programming patterns.

Here are short descriptions:

• dups.txt, nhanes_files.txt example text files for use with the shell examples.
• ex_while_read.sh illustrates the while read pattern for looping.
• ex_check_dup_lines.sh illustrates checking for duplicate lines in a file using sort, uniq, wc and an if statement. The if statement syntax here is for bash and may error in other shells, e.g. zsh.
• ex_variable_expansion.sh demonstrates the difference between single and double quotes in terms of the latter allowing variable expansion.
• recs_data.sh downloads the 2009 and 2015 RECS data and associated codebooks.
• cutnames.sh is an executable program for extracting columns from a csv file by name.
• nhanes_demo.sh downloads several cohorts of NHANES data, converts to csv, selects columns, and appends into a single dataset.

These are files to allow you to view notebooks in this repo through [Binder]. You can ignore these.