Tag, Word & Link Cloud

Plugin for Obsidian

With this plugin you can create a tag, link or word cloud in your notes.

To do this create a codeblock with the language set to either tagcloud, wordcloud or linkcloud.

You can configure your cloud using a YAML syntax.

Tag Cloud

Displays all tags that fit your selection.

Examples

Show all tags in the entire vault

```tagcloud
```

Show all tags in the current file

```tagcloud
source: file
```

Show all tags from a folder/file.

⚠️ Requires Dataview

```tagcloud
source: query
query: Folder/File
```

Show all tags that show up together with our tag.

⚠️ Requires Dataview

```tagcloud
source: query
query: '#yourTag'
```

Show all tags from notes that link to note.

⚠️ Requires Dataview

```tagcloud
source: query
query: '[[Other note]]'
```

Options

Name	Description	Possible Values
query	Requires Dataview, requires source to be set to query	A valid Dataview Source
exclude	exclude certain tags from the tagcloud	an array of tags

All other options from here still apply

Word Cloud

a wordcloud displays all words in your vault/note.

⚠ Word distribution will only be calculated when loading a vault and by running the Recalculate Word Distribution command.

This is because the calculation is computationally expensive and takes some time.¹

Examples

Show all words in the entire vault

```wordcloud
```

Show all words in the current file

```wordcloud
source: file
```

Show all words using a folder/file

```wordcloud
source: file
query: 'Folder/File'
```

Options

Name	Description	Possible Values	Default
stopwords	Remove all stopwords from the result	true/ false	true

All other options from here still apply

Link Cloud

A link cloud displays all links in your vault.

This cloud can only be generated vault wide.

Examples

Show all links

```linkcloud
```

Show all links to existing files

```linkcloud
type: resolved
```

Show all links to non-existing files

```linkcloud
type: unresolved
```

Options

Name	Description	Possible Values	Default
type	Which type of links to show	resolved, unresolved, both	resolved

The following options also apply.

General Options

The following options are supported for all clouds.

Name	Description	Possible Values	Default
shape	What shape to draw	circle, cardioid, diamond, square, triangle-forward, triangle, pentagon, star	circle
source	where are the tags/words coming from?	file, vault, query(only supported in tagcloud)	vault
weight	factor by wich the size of a word is multiplied	any positive integer	2
shrinkToFit	Adjust word weight to make it fit	true/false	true
minCount	Minumum number of occurances	any positive integer	0
maxDepth	Only show the X most used elements(if two elements have the same number of occurrences only one will be counted)	any positive integer (increasing this number may result in the cloud not showing, as only so many elements can be rendered)	25
background	Background color	a hexadecimal RGB value	background color from the chosen theme
width	Width of canvas	in pixels (the px is omitted)	line width
height	Height of canvas	in pixels (the px is omitted)	width / 2
fontFamily	font used to display	any valid font-family
fontWeight	font weight	normal, bold, or a number	normal
minFontSize	minumum font size	any number	0
minRotation	the minimum rotation the text should rotate	number (in rad)	- pi / 2
maxRotation	the maximum rotation the text should rotate	number (in rad)	pi / 2
ellipticity	degree of 'flatness'	number	0.65
shuffle	produce a different looking result each time?	true/false	true
rotateRatio	Rotation Probability	Number as percentage (so 1.0 is 100%)	0.1

Known issues

• In some specific scenario the calculated width off the used canvas element is 0. The plugin will fall back to a value of 500, which depending on the size of your obsidian window, might look strange.

---

Credits

• Built using wordcloud2.js
• and stopword
• as well as the Dataview API
• With help from the code of Tag Wrangler
• Some Code taken from the Word Count Dashboard for Dataview

Footnotes

1. On a high-powered computer with a Vault that contains ~12.000 this takes 15 minutes. ↩