31:T4cb,html`

.debug()

sometimes you just wanna know what the heck is going on. Luckily there's a pretty useful method for doing this- .debug().

This method attempts to reproduce itself on client-side too.

.debug() accepts an optional string, to configure what is shown. Supported formats are 'chunks' and 'highlight'

These methods can be configured in the World object by changing methods available on nlp.world().methods.one.debug.

nlp.verbose()

Additionaly, there's a helpful method - nlp.verbose() to understand what happens doing the POS-tagging steps, which can be really useful to debug where and when a tagging mistake occurrs:

.debug() on the client side:

`32:[] 2f:{"title":"Compromise-output","cells":[{"id":0,"value":"md`# Compromise-output\n\n[compromise](https://beta.observablehq.com/@spencermountain/nlp-compromise) has a bunch of output formats, to allow exporting any kind of data in a reasonable and useful way.\n\nThe are three (main) ways to get data out of a parsed compromise document: \n- **[.text()](https://observablehq.com/@spencermountain/compromise-text)**\n- **[.json()](https://observablehq.com/@spencermountain/compromise-json)**\n- and this method, **.out()**:\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":151,"value":"md`---\n### **.text( {options} )**: - [Docs](https://observablehq.com/@spencermountain/compromise-text)\nreturn a string of text, according to given parameters.\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":39,"value":"nlp('The children are right to laugh at you, Rälf. ').text()","pinned":true,"hidden":false,"mode":"ojs"},{"id":175,"value":"md`---\n### **.json( {options} )**: - [Docs](https://observablehq.com/@spencermountain/compromise-json)\nget structured meta-data from your document\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":178,"value":"nlp(`Wayne's World, party-time, excellent!! 🎸`).json({offset:true})[0]","pinned":true,"hidden":false,"mode":"ojs"},{"id":186,"value":"md`\n---\n## **.out( {options} )**:\nsome popular named output formats:\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":57,"value":"md`\n**.out('tags')** gives you a cleaner output of part-of-speech tags for each term:\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":62,"value":"nlp('powerful like a gorilla, yet soft and yielding like a nerf ball').match('#Adjective').out('tags')","pinned":true,"hidden":false,"mode":"ojs"},{"id":91,"value":"html`

\n sometimes, it's handy to know where you are in the document, using character-offsets.\n

\n .out('offsets') gives you word-start and word-end indexes in the document.\n see full documentation\n

`","pinned":false,"hidden":false,"mode":"ojs"},{"id":93,"value":"{\n let doc = nlp('Goodbye my Coney Island baby')\n return doc.match('coney island').out('offsets')\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":122,"value":"html`

\n .out('freq') is a useful way to count repeating terms, and sort them by frequency:\n

`","pinned":false,"hidden":false,"mode":"ojs"},{"id":124,"value":"{\n let doc = nlp(`What'd I say? Monorail. What's it called? Monorail.`)\n return doc.terms().out('freq')\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":205,"value":"md`---\n## **.html( {segments}, {options} )**:\n\ngenerate html *[or other markup](https://observablehq.com/@spencermountain/compromise-offsets)* from your document, with certain matches wrapped in tags:\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":198,"value":"{\n let doc = nlp(`The aurora borealis? At this time of year?`).debug()\n let str= doc.html({\n '.red': 'aurora #Noun', //a class name : match\n })\n return html`${str}`\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":99,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"id":41,"value":"nlp = await require('compromise')","pinned":false,"hidden":false,"mode":"ojs"},{"id":222,"value":"nlp.version","pinned":false,"hidden":false,"mode":"ojs"},{"id":4,"value":"style = html`