Package 'cat.llm'

Title: LLM-Powered Text Classification for Social Science
Description: Meta-package that installs and loads the full CatLLM ecosystem for LLM-powered text classification across domains. Installing cat.llm brings in cat.stack (general-purpose engine), cat.survey (survey responses), cat.vader (social media), cat.ademic (academic papers), cat.cog (cognitive assessment), cat.pol (policy documents), and cat.web (web content). Like the tidyverse, you can install the full ecosystem or individual domain packages.
Authors: Chris Soria [aut, cre]
Maintainer: Chris Soria <[email protected]>
License: GPL (>= 3)
Version: 3.1.1
Built: 2026-06-04 16:59:37 UTC
Source: https://github.com/chrissoria/cat-llm

Help Index


Attach all CatLLM ecosystem packages

Description

Explicitly loads all domain packages. Normally this happens automatically when library(cat.llm) is called, but this function can be used to reload after detaching.

Usage

catllm_attach()

Value

Invisibly returns a character vector of attached package names.

Examples

## Not run: 
# Normally this happens automatically on `library(cat.llm)`.
# Call manually to re-attach after detaching:
catllm_attach()

## End(Not run)

Domain-suffixed aliases for the CatLLM ecosystem

Description

These functions provide convenient domain-suffixed names so users can tab-complete to find the right function. Each is a thin re-export from the corresponding domain package.

Usage

classify_survey(...)

extract_survey(...)

explore_survey(...)

classify_social(...)

extract_social(...)

explore_social(...)

classify_academic(...)

extract_academic(...)

explore_academic(...)

summarize_academic(...)

cerad_drawn_score(...)

classify_political(...)

extract_political(...)

explore_political(...)

summarize_political(...)

classify_web(...)

extract_web(...)

explore_web(...)

summarize_web(...)

Arguments

...

Additional arguments passed to the Python function.

Examples

## Not run: 
library(cat.llm)

# Survey classification (re-export of cat.survey::classify)
classify_survey(
  input_data      = df$responses,
  categories      = c("Cost", "Quality", "Service", "Other"),
  survey_question = "Why did you choose us?",
  api_key         = Sys.getenv("OPENAI_API_KEY")
)

# Political documents (re-export of cat.pol::classify)
classify_political(
  source     = "city_san_diego",
  doc_type   = "ordinance",
  n          = 50L,
  categories = c("Housing", "Public Safety", "Finance"),
  api_key    = Sys.getenv("OPENAI_API_KEY")
)

# Web content (re-export of cat.web::classify)
classify_web(
  input_data    = c("https://example.com/article-1",
                    "https://example.com/article-2"),
  categories    = c("News", "Opinion", "Tutorial"),
  source_domain = "example.com",
  api_key       = Sys.getenv("OPENAI_API_KEY")
)

# CERAD cognitive scoring (re-export of cat.cog::cerad_drawn_score)
cerad_drawn_score(
  shape       = "circle",
  image_input = "./drawings/",
  api_key     = Sys.getenv("OPENAI_API_KEY")
)

## End(Not run)