Traces: what kind of chart it is
A trace is one series of data, and its type decides what the chart looks like. A dataset with two time-series traces is a chart with two lines; one with a pie trace is a pie.
PlotlyDataSet ds = new PlotlyDataSet();
PlTimeSeriesTrace sales = ds.addTimeSeries("Sales");
sales.mode(TraceMode.MarkersAndLines);
sales.line().shape(PlShape.Spline).smoothing(2.0);
Time series
addTimeSeries(String name) returns a PlTimeSeriesTrace, whose points are a
date and a value.
| Method | What it does |
|---|---|
add(Date, double) / add(Date, double, String text) |
one point, with an optional label |
add(long millis, double, String) |
the same from a timestamp |
mode(TraceMode) |
Lines, Markers, MarkersAndLines, ... |
date() / dateTime() |
whether the x value is a day or a moment |
fill(PlFillType) |
fill the area under the line |
line() |
the line's own settings: shape(PlShape.Spline), smoothing(), width, colour |
type(TraceType) |
the underlying Plotly type, when the default is not wanted |
Bars and named categories
addLabeledSeries(String name) returns a PlLabelValueTrace: the same idea with
a label instead of a date, which is what a bar chart is.
| Method | What it does |
|---|---|
add(String label, double value) |
one bar, or one point |
mode(TraceMode) |
as above |
type(TraceType) |
Bar for bars, the scatter types for points |
Several such traces on one dataset become several bars per category; what happens
to them then is the dataset's barMode(PlBarMode) - side by side, or stacked.
Pies and donuts
addPie() returns a PlPieTrace, filled with label/value pairs.
| Method | What it does |
|---|---|
add(String label, double value) |
one slice |
hole(double) |
0 is a pie; 0.4 is a donut |
textPosition(PlTextPosition) |
labels Inside, Outside, Auto, None |
textInfo(PlTextInfo...) |
what the label says: the name, the value, the percentage |
insideTextFont() / outsideTextFont() |
the fonts of those two |
autoMargin() |
let Plotly make room for labels that stick out |
domain(int x, int y) |
which cell of the dataset's grid this pie goes in |
domain() with the dataset's grid(columns, rows) is how several pies end up
next to each other on one graph.
Sunbursts
addSunburst() returns a PlSunBurstTrace: a hierarchy drawn as rings, where
each item names its parent.
| Method | What it does |
|---|---|
add(String id, String parentId, String label, double value) |
one item, under that parent |
add(String id, String parentId, String label) |
...whose value is the sum of its children |
branchValues(PlBranchValues) |
whether a branch's value is its own or the total of its children |
maxDepth(int) |
how many rings to show at once |
leafOpacity(double) |
how solid the outer ring is |
textPosition(), textInfo(), insideTextOrientation() |
as on the pie |
The root is the item whose parent id is empty.
Gauges
A gauge shows one number on a dial, so it is not added to a dataset like the
others: GaugeDataSource is both the trace and the IPlotlyDataSource, and
is handed to the graph directly.
GaugeDataSource speed = new GaugeDataSource(420, "Speed");
speed.gauge().axis().range(0, 500).tickColor("darkblue");
speed.gauge().step(0, 250, "cyan").step(250, 400, "royalblue");
speed.gauge().threshold().value(490).line().color("red").width(4);
graph.setSource(speed);
| Method | What it does |
|---|---|
value(double) |
the number the needle points at |
mode(PlIndicatorMode...) |
Gauge, Number, Delta - and combinations |
title() |
the text over the dial |
gauge() |
the dial itself: axis() for its range and ticks, bar() for the needle, step(from, to, colour) for its bands, threshold() for the red line |
dataSet() |
the dataset underneath, for a title or a margin around the dial |
The constructors do the common cases: GaugeDataSource(value) and
GaugeDataSource(value, title). The mode is set to gauge-plus-number when the
dataset is asked for, so the number is printed under the dial by default.