This page shows the Week 4 loop so you can see how the pieces fit
together. Your graded files live in your private repo, in the folder
week04-report/.
A Python script reads your CSV and writes two files:
output/numbers.typ (the results sentence and the table)
and output/figure.png (the graph).
You then run typst compile main.typ, and Typst turns
main.typ into output/main.pdf.
The preview lower on this page runs in the browser so you can try
the idea in class. The work you turn in is the PDF you compile on
your own computer.
Install Typst before lab if you can.
week04-report/README.md
is the short lab note you paste on Canvas. You write the prediction,
what you changed, and anything a reader should know about the data.
The script fills in the results.
report/scripts/summarize.py
reads the CSV, writes
output/numbers.typ, and draws
output/figure.png.
The script uses matplotlib. Install it inside this project's
.venv folder so it stays with the lab.
report/main.typ
pulls in the generated snippet (the table and the figure) with
#include.
Run
typst compile main.typ output/main.pdf
to make the paper. Turn in that PDF along with the source files.
Copy the starter from
report/
into a folder named week04-report/ in your private repo.
Run the first build on sample-rt.csv so you can see that
the compiler works. Then point the script at your Week 2 or
Week 3 CSV. In the VS Code terminal:
cd week04-report python3 -m venv .venv source .venv/bin/activate bash build.sh sample-rt.csv # then, on your own file: bash build.sh ../week02-rt/YOURFILE.csv
Windows (PowerShell): py -3 -m venv .venv, then
.\.venv\Scripts\Activate.ps1, then
py -3 -m pip install -r requirements.txt, then
py -3 scripts/summarize.py --csv YOURFILE.csv --out output --report README.md,
then typst compile main.typ output/main.pdf.
If Git is installed, bash build.sh YOURFILE.csv
also works.
On a Mac, install Typst with brew install typst.
On Windows, use winget install --id Typst.Typst.
Check that it worked with typst --version.
The longer note is on workflow.html#typst.
The first code box is the results snippet the script would write
(output/numbers.typ).
The second box is the whole paper
(main.typ): the title, the prediction, the included
results, and a limitation.
The page on the right is a browser preview of how that paper looks.
Compile on your laptop when you want the real PDF.
Load a file, write a prediction, then rebuild the preview.
output/numbers.typRun the preview to fill this block.
main.typThe compiled PDF includes the title, the prediction, the results, and the limitation.
After the first PDF, rewrite the title, prediction, and
limitation in main.typ until another student can
say your claim in one sentence.
Then change something the script controls: which trials you leave
out, how the figure is drawn (how the bars are grouped, or the
first half of the trials versus the second half), or a second
summary number.
Run the build again so output/main.pdf updates.
In the README, quote the old results sentence and the new one.
Keep the mean the script calculated.
Turn in the paths to week04-report/README.md,
week04-report/main.typ,
week04-report/output/numbers.typ,
week04-report/output/figure.png,
and week04-report/output/main.pdf,
plus the path to the CSV you used.
Also name one change you made after the first build, and say what
changed in the PDF. Due Tuesday 29 September. There is no class on the 30th.
build.sh on a sample CSV, opens the PDF, changes one line, and compiles again.Report: the file paths, the CSV you used, and one change you can describe. The mean in the paper comes from the script.