Work with contingency tables

[mbcann01]

Overview

Need to come up with a way to work with 2x2 contingency tables.

Here's what I want to do:

• Input a contingency table
• Display the cases
• Convert from raw data to data frame
• Display contingency table as a data frame
• Display the contingency table as a 2x2 table (matrix)

Helpful websites:

• R Notes notes on making contingency tables
• Check out PercTable as a potential format for the results.

Thoughts

• Enter cross tab as a freq_table.
• Create functions for risks, odds, effect modification, etc.
• Create a function that will allow freq_table to output in table format.
• Create a function that will allow freq_table to output a data frame of observations/cases
• Instead of row and column, consider using exposure and outcome.
• Update vignettes

Other thoughts
freq_crosstab:

• freqtable input
• Manual frequency table inout
• Data frame of observations input
• Output is a table or matrix
• From there, we can calculate quantities of interest (e.g. odds ratio). Broom style.
• Also, need a display only version that can show n’s, row/column percents, etc.

Potential output style

..1	..2	..3	..4
		Col var
Row var	Stats	Outcome +	Outcome -
Exposure +	N
Exposure +	Percent
Exposure -	N
Exposure -	Percent

Maybe even display as a flextable. Document that this layout is for viewing rather than analysis?
Maybe just use gmodels::CrossTable() instead of reinventing the wheel?

• Make a freqtable_to_matrix function. https://brad-cannell.github.io/r_notes/contingency-tables.html#convert-a-frequency-table-into-a-contingency-table.
• Make a freqtable_to_df function. https://brad-cannell.github.io/r_notes/contingency-tables.html#convert-a-frequency-table-into-a-data-frame-of-observations.