Accesses the Monkeylearn API for Text Classifiers and Extractors

Allows using some services of Monkeylearn <> which is a Machine Learning platform on the cloud for text analysis (classification and extraction).


monkeylearn 0.2.0

  • New functions monkey_classify() and monkey_extract() that:

    • Accept as input both a vector and a dataframe and named column
    • Always return a tibble explicitly relating each input to its classification, allowing for the removal of the MD5 hash
    • Have an unnest flag to unnest the output (turn 1 row per input into 1 row per output)
    • Have a .keep_all flag to retain other columns if input is a dataframe
    • Coerce NULL values and empty vectors returned from MonkeyLearn to NAs
    • Include inputs that could not be processed as NAs in the output
    • Message the first 20 indices of inputs that are not sent to the API (these now include NA and NULL values as well as empty strings)
    • Message the currently processing batch
  • Bug fixes and improvements to monkeylearn_classify() and monkeylearn_extract()

    • monkeylearn_classify() can now accept params
    • Fix to messaging when unable to connect to MonkeyLearn API
    • Default texts per request is set to 200 now (the recommended number), rather than 20
    • Addition of message suggesting that users switch to newer functions
  • Implementation of ratelimitr. Creation and documentation of two environment variables allowing smarter rate handling when querying the MonkeyLearn API.

  • Creation of pkgdown website

  • Programmatic test coverage to re-use common tests for multiple circumstances.

  • Use of a cowsay monkey when verbose=TRUE.

monkeylearn 0.1.3

  • Better states the dependency on tibble, it is tibble >= 1.2.

  • Better handles blank text in input, outputs an empty tibble and a warning if the request is only blank, and a message if only parts of the request are blank.

monkeylearn 0.1.2

  • Disables HTTP2 for now because of a bug for Windows users. Fix by Jeroen Ooms.

monkeylearn 0.1.1

  • Added a file to track changes to the package.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


0.2.0 by Maëlle Salmon, 9 months ago,

Report a bug at

Browse source code at

Authors: Maëlle Salmon [aut, cre] , Amanda Dobbyn [aut] , Thomas Leeper [rev] (Thomas Leeper reviewed the package for rOpenSci , see , Jeroen Ooms [ctb] , rOpenSci [fnd] ( , Earlybird Software [fnd] (

Documentation:   PDF Manual  

Task views: Natural Language Processing

GPL (>= 2) license

Imports cowsay, digest, dplyr, httr, jsonlite, magrittr, methods, purrr, tibble, tidyr, ratelimitr

Suggests knitr, rmarkdown, testthat

See at CRAN