There are two tpyo's in the CSV column headers that I've corrected as I keep on typing the correct spelling and getting
errors.
rapid_enlargment > rapid_enlargement
compressive_symtoms > compressive_symptoms
@mdp21oe : You should correct these in the copy of the data/csv/sheffield_thyroid_nodule.csv so that the script will run correctly.
Having subsections in the sections/ directory was causing problems when rendering as Quarto treats each child .qmd as an individual "Notebook" and attempts to render it when building a manuscript. These Notebooks couldn't access dataframes or objects created in index.qmd.
The solution was to prefix each with an _ which forces them to be ignored as Notebooks. Each is then included verbatim via the {{< include sections/_file.qmd >}} lines from index.qmd.
I've also moved the setup and recipe code chunk to their own child documents and collated all of the logistic regression sections into their own child document which is included under the results section (for now).
I've also split the logistic regression into its own section and rather than the manual extraction of coefficients and confidence intervals and exponentiating them and combining them into a single table I've used gtsummary::tbl_regression() to summarise the coefficients. This swaps one set of fiddly writing for another but is perhaps a little easier to work with.
There are two tpyo's in the CSV column headers that I've corrected as I keep on typing the correct spelling and getting errors.
rapid_enlargment>rapid_enlargementcompressive_symtoms>compressive_symptoms@mdp21oe : You should correct these in the copy of the
data/csv/sheffield_thyroid_nodule.csvso that the script will run correctly.Having subsections in the
sections/directory was causing problems when rendering as Quarto treats each child.qmdas an individual "Notebook" and attempts to render it when building a manuscript. These Notebooks couldn't access dataframes or objects created inindex.qmd.The solution was to prefix each with an
_which forces them to be ignored as Notebooks. Each is then included verbatim via the{{< include sections/_file.qmd >}}lines fromindex.qmd.I've also moved the
setupandrecipecode chunk to their own child documents and collated all of the logistic regression sections into their own child document which is included under the results section (for now).I've also split the logistic regression into its own section and rather than the manual extraction of coefficients and confidence intervals and exponentiating them and combining them into a single table I've used
gtsummary::tbl_regression()to summarise the coefficients. This swaps one set of fiddly writing for another but is perhaps a little easier to work with.