Describe a data frame for AI context

Creates a formatted description of a data frame suitable for AI context. Can use different methods to generate the description.

Usage

cassidy_describe_df(
  data,
  name = NULL,
  method = c("codebook", "skim", "basic"),
  include_summary = TRUE,
  max_vars = NULL,
  show_sample = TRUE
)

Arguments

data

A data frame or tibble

name

Optional character string to use as the data frame name. If NULL, attempts to use the name of the object passed to data.

method

Method to use: "codebook" (default), "skim", or "basic"

include_summary

Include statistical summaries (for "basic" method)

max_vars

Maximum variables to show (NULL = all)

show_sample

Include sample values? (Default: TRUE)

Value

An object of class cassidy_df_description containing the formatted description text

Details

The method parameter controls how the data frame is described:

• "codebook": Uses cassidy_describe_codebook() which provides a compact, markdown-formatted description optimized for LLM consumption, including variable labels, value labels, factor levels, and more

• "skim": Uses skimr::skim() if available, which provides summary statistics organized by variable type. Falls back to "basic" if not installed.

• "basic": Uses base R to provide a simple description with variable types and basic summaries.

Examples

if (FALSE) { # \dontrun{
  # Using codebook method (most detailed)
  desc <- cassidy_describe_df(mtcars, method = "codebook")

  # With explicit name
  desc <- cassidy_describe_df(my_data, name = "survey_responses", method = "skim")

  # Using basic method (lightweight)
  desc <- cassidy_describe_df(mtcars, method = "basic")

  # Use in chat
  cassidy_chat("What analyses would you recommend?", context = desc)
} # }