Demo: nicgirault.github.io/circosJS/demo
Circos is a javascript library to easily build interactive graphs in a circular layout. It's based on d3.js. It aims to be a javascript version of the Circos software.
You should consider using Circos to show:
- relationships between entities
- periodical data
Average temperatures in Paris from 2007 (inner) to 2014 (outer). The circular layout highlights seasonal effect.
If you don't know what is yarn or npm you can skip this step and get started with this canvas. Otherwise:
yarn add circos
To instantiate a new circos:
var myCircos = new Circos({
container: '#chart',
width: 500,
height: 500,
});
A circos graph is based on a circular axis layout. Data tracks appear inside and/or outside the circular layout.
In order to place data on the circos graph, you must first specify the layout.
myCircos.layout(data, configuration);
The first argument of the layout
function is a configuration object that control the format of the layout.
Here are the default parameters for a layout:
var configuration = {
innerRadius: 250,
outerRadius: 300,
cornerRadius: 10,
gap: 0.04, // in radian
labels: {
display: true,
position: 'center',
size: '14px',
color: '#000000',
radialOffset: 20,
},
ticks: {
display: true,
color: 'grey',
spacing: 10000000,
labels: true,
labelSpacing: 10,
labelSuffix: 'Mb',
labelDenominator: 1000000,
labelDisplay0: true,
labelSize: '10px',
labelColor: '#000000',
labelFont: 'default',
majorSpacing: 5,
size: {
minor: 2,
major: 5,
}
},
clickCallback: null
}
The second argument of the layout
function is an array of data that describe the layout regions. Each layout region must have an id and a length. You can also specify a color and a label.
var data = [
{ len: 31, color: "#8dd3c7", label: "January", id: "january" },
{ len: 28, color: "#ffffb3", label: "February", id: "february" },
{ len: 31, color: "#bebada", label: "March", id: "march" },
{ len: 30, color: "#fb8072", label: "April", id: "april" },
{ len: 31, color: "#80b1d3", label: "May", id: "may" },
{ len: 30, color: "#fdb462", label: "June", id: "june" },
{ len: 31, color: "#b3de69", label: "July", id: "july" },
{ len: 31, color: "#fccde5", label: "August", id: "august" },
{ len: 30, color: "#d9d9d9", label: "September", id: "september" },
{ len: 31, color: "#bc80bd", label: "October", id: "october" },
{ len: 30, color: "#ccebc5", label: "November", id: "november" },
{ len: 31, color: "#ffed6f", label: "December", id: "december" }
]
The id
parameter will be used to place data points on the layout.
To visualize the result:
myCircos.render();
A track is a series of data points.
To add a track to your graph you should write something like this:
myCircos.heatmap(
'my-heatmap',
data,
{
// your heatmap configuration (optional)
},
);
This pattern is similar to all track types:
myCircos.trackType('track-id', data, configuration);
Note: The track name is used as a HTML class name so here are the format limitations.
- Must be unique.
- Should be slug style for simplicity, consistency and compatibility. Example:
my-heatmap
- Lowercase, a-z, can contain digits, 0-9, can contain dash or dot but not start/end with them.
- Consecutive dashes or dots not allowed.
- 50 characters or less.
Chords tracks connect layout regions.
Some gene fusions in human karyotype (source)
Data should looks like this:
var data = [
// sourceId, sourceStart, sourceEnd, targetId, targetStart, targetEnd
['january', 1, 12, 'april', 18, 20],
['february', 20, 28, 'december', 1, 13],
];
Optionally each datum can define a seventh element which can be used to be interpreted as a value
to draw colored ribbons with palettes or a color function.
The default configuration is:
{
color: '#fd6a62',
opacity: 0.7,
zIndex: 1,
tooltipContent: null,
min: null,
max: null,
logScale: false,
logScaleBase: Math.E,
}
Electrical comsumption in France in 2014
To add a heatmap to your circos instance:
myCircos.heatmap('electrical-consumption', data, configuration);
Configuration:
{
innerRadius: null,
outerRadius: null,
min: null,
max: null,
color: 'YlGnBu',
logScale: false,
tooltipContent: null,
}
Data format:
var data = [
// each datum should be
// layout_block_id, start, end, value
['january', 0, 1, 1368001],
['january', 1, 2, 1458583],
['january', 2, 3, 1481633],
['january', 3, 4, 1408424]
...
['february', 0, 1, 1577419],
['february', 1, 2, 1509311],
['february', 2, 3, 1688266],
...
]
Human karyotype with cytobands highlighted (source)
To add a highlight to your circos instance:
myCircos.highlight('cytobands', data, configuration);
The minimal datum should have block_id
, start
and end
attributes.
Configuration:
{
innerRadius: null,
outerRadius: null,
min: null,
max: null,
color: 'd3d3d3',
strokeColor: null,
strokeWidth: 0,
opacity: 1,
logScale: false,
tooltipContent: null,
}
Genetic abnormalities in human stem cells (source)
Data should looks like this:
var data = [
{
block_id: 'january',
start: 1,
end: 10,
value: 5
}
];
The available configuration fields are:
Some single nucleotide polymorphism on chromosomes 1, 2 and 3 (source)
myCircos.line('line1', data, configuration);
The minimal datum should have block_id
, position
and value
attributes.
The available configuration fields are:
- innerRadius
- outerRadius
- color
- strokeColor
- strokeWidth
- direction
- fill
- maxGap
- opacity
- zIndex
- min
- max
- logScale
- logScaleBase
- axes
- backgrounds
Note: The tooltip option is not available for line track. To display a tooltip, you should superimpose an invisble scatter
track (fill: false
and strokeWidth: 0
) and set a tooltip for this track.
myCircos.scatter('scatter1', data, configuration);
The minimal datum should have block_id
, position
and value
attributes.
The available configuration fields are:
- innerRadius
- outerRadius
- color
- strokeColor
- strokeWidth
- direction
- fill
- size
- shape
- opacity
- zIndex
- min
- max
- logScale
- logScaleBase
- axes
- backgrounds
myCircos.stack('stack', data, configuration);
The minimal datum should have block_id
, start
and end
attributes.
Configuration:
{
innerRadius: null,
outerRadius: null,
min: null,
max: null,
color: '#fd6a62',
strokeColor: '#d3d3d3',
strokeWidth: 1,
direction: 'out',
thickness: 10,
radialMargin: 2,
margin: 2,
opacity: 1,
logScale: false,
tooltipContent: null,
}
myCircos.text('text', data, configuration);
The minimal datum should have block_id
, position
and value
attributes.
Configuration:
{
innerRadius: null,
outerRadius: null,
style: {
'font-size': 12,
fill: 'black',
opacity: 1,
},
}
You can add a list of backgrounds:
{
backgrounds: [
{
start: 0.006,
color: '#4caf50',
opacity: 0.1
},
{
start: 0.002,
end: 0.006,
color: '#d3d3d3',
opacity: 0.1
},
{
end: 0.002,
color: '#f44336',
opacity: 0.1
}
]
}
The start
and end
fields are interpreted as values on the same scale than the track values.
- If
start
is not specified, default is themin
value of the track. - If
end
is not specified, default is themax
value of the track.
You can also specify a color
and an opacity
.
For the layout, the innerRadius and outerRadius values are always interpreted as a number of pixel.
For tracks:
If innerRadius and outerRadius are between 0
and 1
, the value is interpreted as a fraction of the innerRadius of the layout.
eg:
{
innerRadius: 0.5,
outerRadius: 0.8
}
If innerRadius and outerRadius are between 1
and 10
, the value is interpreted as a fraction of the outerRadius of the layout.
eg:
{
innerRadius: 1,
outerRadius: 1.2
}
Otherwise it is interpreted as a number of pixels.
The default min and max values are computed according to the dataset. You can override these values by specifying a min
or max
attribute in the configuration.
The color attribute can be either:
e.g #d3d3d3
, blue
, rgb(0, 0, 0)
Palette name from d3-scale-chromatic
The color will be computed dynamically according to the track data value
field.
If you prefix the palette name with a -
(e.g -BrBG
), the palette will be reversed.
The list of palette names are:
BrBG: PRGn: PiYG: PuOr: RdBu: RdGy: RdYlBu: RdYlGn: Spectral: Blues: Greens: Greys: Oranges: Purples: Reds: BuGn: BuPu: GnBu: OrRd: PuBuGn: PuBu: PuRd: RdPu: YlGnBu: YlGn: YlOrBr: YlOrRd:
You can specify a function that compute the color code given the track data and the datum index. For example:
{
color: function(datum, index) {
return datum.value < 5 ? 'red' : 'green'
}
}
The default value is an empty array:
{
axes: []
}
You can add items to this array to render an axis or a group of axes. You can give axes a color
(default: '#d3d3d3'), thickness
(default: 1) and opacity
(default: track opacity):
{
axes: [
{
color: 'black',
thickness: 2, // in pixel
opacity: 0.3 // between 0 and 1
}
]
}
Then you should specify where to place the axes.
You can either define a single axis by defining a position
attribute with a value between the min and max value of the track:
{
axes: [
{
color: 'red',
position: 4
},
{
color: 'green',
position: 15
}
]
}
Or define a range of axes with a spacing
attribute and optionnally a start
and end
attributes:
{
axes: [
{
spacing: 2
}
]
}
Here is an advanced example:
{
axes: [
{
color: 'red',
spacing: 2,
end: 4
},
{
color: 'green',
spacing: 2,
start: 16
},
{
spacing: 2,
start: 4,
end: 16,
thickness: 2
},
{
spacing: 1,
start: 4,
end: 16,
thickness: 1
}
]
}
The values that you set for position
, spacing
, start
and end
are in the unit of the track values.
A function that receive the datum and the index as a value and return a string displayed in the tooltip (HTML is accepted):
{
tooltipContent: function (datum, index) {
return `<h5>${datum.block_id}:${datum.start}-${datum.end} ➤ ${datum.value}</h5> <i>(CTRL+C to copy to clipboard)</i>`
}
}
Then when you mouseover the datum, a tooltip will be displayed. Note that you can use the keyboard shortcut CTRL+C to copy the content to clipboard.
Show or not a tooltip with the value of the axis. Default is true
.
It should be either in
or out
. Default is out
. For stack you can also use center
.
true
or false
.
true
or false
. Default is false
.
The log base if logScale is true
. Default is Math.E
.
It should be one of:
circle
cross
diamond
square
triangle
star
wye
This should be an integer. The higher it is the more above the track will appear.
Nicolas Girault [email protected]
Your feedbacks are welcome. If you're struggling using the librairy, the best way to ask questions is to use the Github issues so that they are shared with everybody.