We Keep Coding

Getting the closest intersection to a click on Chart.js

I'm using chart.js to make a radar chart that the user should be able to interact on their own. The idea is that if the user clicks on the red spot, the app should know whether the green spot is the closest intersection of ticks and angle lines, so that it can change the dataset value at index 1 (corresponding to B) from 2 (corresponding to mid-low) to 4 (mid-high).
And it's set up like this:
const chartData = {
labels: ['A', 'B', 'C', 'D'],
datasets: [{
fill: true,
backgroundColor: 'rgba(6,211,248,0.5)',
borderColor: 'rgb(6,211,248)',
data: mockData.presetValues.a,
}]
};
const chartConfig = {
type: 'radar',
data: chartData,
options: {
elements: {
line: {
borderWidth: 3
},
},
scales: {
r: {
min: 0,
max: 5,
ticks: {
stepSize: 1,
callback: function (value) {
return ['none', 'low', 'mid-low', 'mid', 'mid-high', 'high'][value]
}
},
grid: {
circular: true,
lineWidth: 5
},
angleLines: {
color: 'black',
}
}
},
plugins: {
legend: {
display: false
}
},
events: ['click']
}
};
const actionChart = new Chart(
canvas,
chartConfig
);
Then I have a listener set up for it like so:
canvas.addEventListener('click', handleChartClick)
And finally I have a handler like this, which is where my problem is:
function handleChartClick(ev) {
let clickPosition = Chart.helpers.getRelativePosition(ev, actionChart);
console.log(clickPosition);
//Find closest tick to click and change values in dataset before refresh
}
After hours of scouring the documentation and this site, I can not find a way of correlating the position of the click that I receive from the getRelativePosition method with the position of the intersections between the angle lines and the ticks. Once I have those two pieces of information I can modify the dataset and refresh, but I'd greatly appreciate if anyone could help me figure out how to get the closest intersection to the click.

You could use onClick option of chart in order to catch click event (instead of adding a listener to the canvas). In this way the passed event is already normalized.
The callback could be something like that, in order to have the closest tick:
options: {
...
onClick(event, elements, chart) {
const scale = chart.scales.r;
const posY = Math.abs(scale.getDecimalForPixel(event.y) - 0.5);
const posX = Math.abs(scale.getDecimalForPixel(event.x) - 0.5);
const scalePosition = Math.max(posY, posX);
const value = Math.round(scalePosition * 10);
console.log(['none', 'low', 'mid-low', 'mid', 'mid-high', 'high'][value]); // shows value
},
...

Apexcharts max width on item bars

How can I change the maximum width/height an item bar has in Apexcharts? (https://apexcharts.com/)
I have an horizontal bar chart whose data are loaded dynamically and the number of users (items) varies. When there are few or only one user the bar looks very huge, I would like to set the height to max 50px.
The code:
var timeChart = {
formatter: function (value) {
var v = formatFromSeconds(value);
return v;
}
};
var desv_time = function (value) {
var v = formatFromSeconds(value);
if (value>0)return "+"+v;
else return v;
}
//CHART
var chart1 = new ApexCharts( document.querySelector("#chart-desv"),
{
chart: {
id: 'chart1',
type: 'bar'
},
plotOptions: {
bar: {
horizontal: true,
},
},
dataLabels: {
enabled: true,
formatter: desv_time,
style: {
colors: ["#000"]
}
},
series: [],
xaxis: {labels: {
formatter: desv_time,
}},
tooltip: {
followCursor: true,
x: {
formatter: function(value) {
return value;
},
},
y: {
formatter: desv_time,
title: {formatter: (seriesName) => 'Time difference',}
},
marker: {show:false}
},
title: { text: "Usual workday duration: 08:00", align: "center" }
});
chart1.render();
Screenshot (very huge bar with one user):

I'm using an approach similar to #Oscar Schafer's
I've found that, for my case, the width could be based on the length of my series.
The premises are:
When we have few items, e.g. 1 or 2, the chart width will be filled with only these columns. Then, if I have few items, the columnWidth should be proportionally smaller (close to 20% in my case)
When we have more items, they all share the chart width and fill it more smoothly. Then, if I have a lot of items, the column should be proportionally larger (close to 80%)
I've used the LOGISTIC-GROWTH MODEL as a function to calculate the bar width
Here is the code for to calculate the column width:
// f(x) = c / (1 + a*exp(-x*b)) -> LOGISTIC GROWTH MODEL
var optimalColumnWidthPercent = 20 + (60 / (1 + 30*Math.exp(-seriesLength /3)));
// 20: minimum width should be close to 20 (when only one item)
// 20+60: maximum width should be close 80
// 30 and 3: the a and b from the function, I've selected after testing some cenarios from seriesLength from 1 to 12 and finding which worked for me
...
plotOptions: {
bar: {
columnWidth: optimalColumnWidthPercent + "%"
},
},

For a horizontal bar chart, you can use the barHeight property to set the height of the bars. Unfortunately it can't be set as a pixel value but instead is "the percentage of the available height in the grid-rect" (from docs).
Effectively it works as a maximum height, so the bars will shrink if there's more data that needs to fit.
plotOptions: {
bar: {
barHeight: '70%'
}
}

Here's how I achieve this behaviour in VueJS.
Template part
<apexchart :options="barChartOptions" :series="series"></apexchart>
Script Part
computed() {
chartOptions() {
return {
chart: {
type: 'bar',
height: 100 + this.myData.length() * barWidth,
}
}
}
}
Where myData is an array of data that contains the information to be plotted and barWidth is the width of each bar that I want.
Note that this essentially fixes the width of each bar and then resizes the height of the chart as a linear function of the number of bars.

For a horizontal bar chart, you can use the columnWidth property to set the widthof the bars. Unfortunately it can't be set as a pixel value but instead is "the percentage of the available widthin the grid-rect" (from docs).
Effectively it works as a maximum width, so the bars will shrink if there's more data that needs to fit.
here,
plotOptions: {
bar: {
horizontal: false,
borderRadius: 10,
columnWidth: '45%',
},
},

Chart.js not showing dynamically populated data

I was having some trouble when trying to dynamically populate bar chart in chart.js. I have two arrays, one for label, one for its price and both of them are already populated with the sorted data from firebase. Here is my code:
var ctx = document.getElementById('brandChart').getContext("2d");
var data = {
labels: [],
datasets: [{
data: [],
backgroundColor: [
"#424242",
]
}]
};
var options = {
layout: {
padding: {
top: 5
}
},
responsive: true,
legend: {
display: true,
position: 'bottom',
// disable legend onclick remove slice
onClick: null
},
animation: {
animateScale: true,
animateRotate: true
},
};
var opt = {
type: "horizontalBar",
data: data,
options: options
};
if (brandChart) brandChart.destroy();
brandChart = new Chart(ctx, opt);
// dynamically populate chart
for(var i = 0; i < labelData.length; i++){
brandChart.config.data.labels.push(labelData[i]);
}
for(var i = 0; i < priceData.length; i++){
brandChart.config.data.datasets[0].data.push(priceData[i]);
}
brandChart.update();
I managed to show all of them in bar chart, however, the result as such:
It is kind of squeeze between each labels if there are too many categories. Also, only the first bar has the color & the legends shown undefined. Any ideas how to solve these?

ɪꜱꜱᴜᴇ #1 - ꜱᴏʟᴜᴛɪᴏɴ
Add a callback for y-axis ticks, in your chart options :
options: {
scales: {
yAxes: [{
ticks: {
callback: function(t, i) {
if (!(i % 2)) return t;
}
}
}]
},
...
}
this will only show every other label on y-axis.
ɪꜱꜱᴜᴇ #2 - ꜱᴏʟᴜᴛɪᴏɴ
This is because, you have only one color in your backgroundColor array. If you want different color for each bar, then you need to populate this array with multiple color values.
Edit: as it seems form your updated question, you already kind of got the idea.
ɪꜱꜱᴜᴇ #3 - ꜱᴏʟᴜᴛɪᴏɴ
Define the label property for your dataset , like so :
datasets: [{
label: 'Legend Title', //<- define this
data: [],
backgroundColor: ["#424242", ]
}]

How to detect when Tooltip closes in chart.js?

I'm using chart.js for a web project and it's working pretty fine. However, I do have one question. I'm trying to connect a line graph with n data points to a list of n html divs. When the user hovers over data point 2, div 2 will be highlighted and a function is called. That does work. However, when the user unhovers data point 2, div 2 should change its style back to the default style.
My question is: How can I detect the mouseout event on data points?
That is how I define what happens when the data point is hovered.
myChart = new Chart(ctx, {
type: 'line',
data: chartData,
options: {
title: {
...
},
tooltips: {
enabled: true,
custom: function(tooltip) {
if (!tooltip) {
return;
}
if(tooltip.dataPoints != null) {
// here, the function that highlights the respective div is called, and it works fine
}
}
}
}
});
Is there such a thing for unhovering? I found out that there is a global events -> mousout option, but I don't figure out how to use it and I also think that it references the whole chart.
Thank you!

Not sure if this will help you, but I had a similar issue with stacked bar charts. I wanted to show values at the top of the bars, but I found that if the tooltips were open the values were written over the top of the tooltips, making both unreadable. I decided I wanted to show the values only if the tooltips were not showing (and were not rendered if a tooltip was open).
Turns out I can use the tooltip's opacity setting to determine if the tooltip is showing or not. This is very over-simplified, but this is what I came up with:
options: {
tooltips: {
custom: function( tooltip ) {
if( tooltip.opacity > 0 ) {
console.log( "Tooltip is showing" );
} else {
console.log( "Tooltip is hidden" );
}
return;
}
}
}
Having worked that out, I was then able to save a global variable that I could test elsewhere to see if the tooltip was showing.

var ctx = document.getElementById("canvas").getContext("2d");
var data = {
labels: [
"Red",
"Green",
"Yellow"
],
datasets: [{
data: [300, 50, 100],
backgroundColor: [
"#FF6384",
"#36A2EB",
"#FFCE56"
],
hoverBackgroundColor: [
"#FF6384",
"#36A2EB",
"#FFCE56"
]
}]
};
Chart.pluginService.register({
beforeRender: function(chart) {
if (chart.config.options.showAllTooltips) {
// create an array of tooltips
// we can't use the chart tooltip because there is only one tooltip per chart
chart.pluginTooltips = [];
chart.config.data.datasets.forEach(function(dataset, i) {
chart.getDatasetMeta(i).data.forEach(function(sector, j) {
chart.pluginTooltips.push(new Chart.Tooltip({
_chart: chart.chart,
_chartInstance: chart,
_data: chart.data,
_options: chart.options.tooltips,
_active: [sector]
}, chart));
});
});
// turn off normal tooltips
chart.options.tooltips.enabled = false;
}
},
afterDraw: function(chart, easing) {
if (chart.config.options.showAllTooltips) {
// we don't want the permanent tooltips to animate, so don't do anything till the animation runs atleast once
if (!chart.allTooltipsOnce) {
if (easing !== 1)
return;
chart.allTooltipsOnce = true;
}
// turn on tooltips
chart.options.tooltips.enabled = true;
Chart.helpers.each(chart.pluginTooltips, function(tooltip) {
tooltip.initialize();
tooltip.update();
// we don't actually need this since we are not animating tooltips
tooltip.pivot();
tooltip.transition(easing).draw();
});
chart.options.tooltips.enabled = false;
}
}
})
var myPieChart = new Chart(ctx, {
type: 'pie',
data: data,
options: {
showAllTooltips: true
}
});
<script src="https://cdn.jsdelivr.net/npm/chart.js#2.8.0/dist/Chart.min.js"></script>
<canvas id="canvas"></canvas>

Chart.js: Bar Chart Click Events

I've just started working with Chart.js, and I am getting very frustrated very quickly. I have my stacked bar chart working, but I can't get the click "events" to work.
I have found a comment on GitHub by nnnick from Chart.js stating to use the function getBarsAtEvent, even though this function cannot be found in the Chart.js documentation at all (go ahead, do a search for it). The documentation does mention the getElementsAtEvent function of the chart reference, but that is for Line Charts only.
I set an event listener (the right way) on my canvas element:
canv.addEventListener('click', handleClick, false);
...yet in my handleClick function, chart.getBarsAtEvent is undefined!
Now, in the Chart.js document, there is a statement about a different way to register the click event for the bar chart. It is much different than nnnick's comment on GitHub from 2 years ago.
In the Global Chart Defaults you can set an onClick function for your chart. I added an onClick function to my chart configuration, and it did nothing...
So, how the heck do I get the on-click-callback to work for my bar chart?!
Any help would be greatly appreciated. Thanks!
P.S.: I am not using the master build from GitHub. I tried, but it kept screaming that require is undefined and I was not ready to include CommonJS just so that I could use this chart library. I would rather write my own dang charts. Instead, I downloaded and am using the Standard Build version that I downloaded straight from the link at the top of the documentation page.
EXAMPLE: Here is an example of the configuration I am using:
var chart_config = {
type: 'bar',
data: {
labels: ['One', 'Two', 'Three'],
datasets: [
{
label: 'Dataset 1',
backgroundColor: '#848484',
data: [4, 2, 6]
},
{
label: 'Dataset 2',
backgroundColor: '#848484',
data: [1, 6, 3]
},
{
label: 'Dataset 3',
backgroundColor: '#848484',
data: [7, 5, 2]
}
]
},
options: {
title: {
display: false,
text: 'Stacked Bars'
},
tooltips: {
mode: 'label'
},
responsive: true,
maintainAspectRatio: false,
scales: {
xAxes: [
{
stacked: true
}
],
yAxes: [
{
stacked: true
}
]
},
onClick: handleClick
}
};

I managed to find the answer to my question by looking through the Chart.js source code.
Provided at line 3727 of Chart.js, Standard Build, is the method .getElementAtEvent. This method returns me the "chart element" that was clicked on. There is sufficent data here to determine what data to show in a drill-down view of the dataset clicked on.
On the first index of the array returned by chart.getElementAtEvent is a value _datasetIndex. This value shows the index of the dataset that was clicked on.
The specific bar that was clicked on, I believe, is noted by the value _index. In my example in my question, _index would point to One in chart_config.data.labels.
My handleClick function now looks like this:
function handleClick(evt)
{
var activeElement = chart.getElementAtEvent(evt);
..where chart is the reference of the chart created by chart.js when doing:
chart = new Chart(canv, chart_config);
The specific set of data that was selected by the click can therefore be found as:
chart_config.data.datasets[activeElement[0]._datasetIndex].data[activeElement[0]._index];
And there you have it. I now have a datapoint that I can build a query from to display the data of the bar that was clicked on.
AUGUST 7TH, 2021. UPDATE
There is now a method for what we are looking for. Take a look at here

Hi this is the click event under options which is getting values from x and y-axis
onClick: function(c,i) {
e = i[0];
console.log(e._index)
var x_value = this.data.labels[e._index];
var y_value = this.data.datasets[0].data[e._index];
console.log(x_value);
console.log(y_value);
}

I found this solution at https://github.com/valor-software/ng2-charts/issues/489
public chartClicked(e: any): void {
if (e.active.length > 0) {
const chart = e.active[0]._chart;
const activePoints = chart.getElementAtEvent(e.event);
if ( activePoints.length > 0) {
// get the internal index of slice in pie chart
const clickedElementIndex = activePoints[0]._index;
const label = chart.data.labels[clickedElementIndex];
// get value by index
const value = chart.data.datasets[0].data[clickedElementIndex];
console.log(clickedElementIndex, label, value)
}
}
}

You can use onClick like this.
var worstCells3GBoxChart = new Chart(ctx, {
type: 'bar',
data: {
labels: lbls,
datasets: [{
label: 'Worst Cells by 3G',
data: datas,
backgroundColor: getColorsUptoArray('bg', datas.length),
borderColor: getColorsUptoArray('br', datas.length),
borderWidth: 1
}]
},
options: {
legend: {
display: false
},
scales: {
yAxes: [{
ticks: {
beginAtZero: true
}
}]
},
onClick: function (e) {
debugger;
var activePointLabel = this.getElementsAtEvent(e)[0]._model.label;
alert(activePointLabel);
}
}
});

Chartjs V3.4.1
This is what worked for me in v3, after looking at solutions for older versions:
const onClick = (event, clickedElements) => {
if (clickedElements.length === 0) return
const { dataIndex, raw } = clickedElements[0].element.$context
const barLabel = event.chart.data.labels[dataIndex]
...
}
raw is the value of the clicked bar.
barLabel is the label of the clicked bar.
You need to pass the onClick to the bar chart config:
const barConfig = {
...
options: {
responsive: true,
onClick,
...
}
}

Well done! This seems to return the data value being charted though, which in many cases might be possible to appear more than once, thus making it unclear what was clicked on.
This will return the actual data label of the bar being clicked on. I found this more useful when drilling down into a category.
chart_config.data.labels[activeElement[0]._index]

I was able to make this work in another way.
Might not be supported, but sometimes, I find that neither the label nor the value is adequate to get me the necessary information to populate a drill-through.
So what I did was add a custom set of attributes to the data:
var ctx = document.getElementById("cnvMyChart").getContext("2d");
if(theChart != null) {theChart.destroy();}
theChart = new Chart(ctx, {
type: typ,
data: {
labels: ["Red", "Blue", "Yellow", "Green", "Purple", "Orange"],
datakeys: ["thefirstone","thesecondone","thethirdone","thefourthone","thefifthone","thesixthone"],
datasets: [{
label: '# of Votes',
data: [12, 19, 3, 5, 2, 3],
...etc
Then when I need to push the drillthrough key into another ajax call, I was able to get it with this:
var theDrillThroughKey = theChart.config.data.datakeys[activePoints[0]._index];
So I'm really not sure that it's appropriate to be adding custom elements into the data for the Chart, but it's working so far in Chrome, IE and Firefox. I needed to be able to put more information into the drillthrough than I really wanted displayed.
Example of the full thing: https://wa.rrdsb.com/chartExamples
Thoughts?

I had the same problem with multiple datasets, and used this workaround:
var clickOnChart = function(dataIndex){
...
}
var lastHoveredIndex = null;
var chart_options = {
...
tooltips: {
...
callbacks: {
label: function(tooltipItem, chart) {
var index = tooltipItem.datasetIndex;
var value = chart.datasets[index].data[0];
var label = chart.datasets[index].label;
lastHoveredIndex = index;
return value + "€";
}
}
},
onClick:function(e, items){
if ( items.length == 0 ) return; //Clicked outside any bar.
clickOnChart(lastHoveredIndex);
}
}

Let's say that you declared a chart using a method like so:
window.myBar = new Chart({chart_name}, {
type: xxx,
data: xxx,
events: ["click"],
options: {
...
}
});
A good way of declaring onclick events would involve listening for the canvas click, like so:
({chart_name}.canvas).onclick = function(evt) {
var activePoints = myBar.getElementsAtEvent(evt);
// let's say you wanted to perform different actions based on label selected
if (activePoints[0]._model.label == "label you are looking for") { ... }
}

In the chart options for Chart.js v3.5.1 which is latest
Check below sample code
let enterpriseChartOptions = {
responsive:true,
maintainAspectRatio: false,
onClick: (c,i) => {
console.log('Get the underlying label for click,', c.chart.config._config.data.labels[i[0].index]);
},
plugins: {
title:{
text:'Enterprise Dashboard (Health Status of 10 stores) updated every 30 minutes',
fontSize:20
},
},
scales: {
x: {
display: true,
type: 'category',
position: 'right',
ticks: {
padding: 8,
},
},
y: {
display: true,
ticks: {
callback: function(val, index) {
// Show the label
return val < 1 ? "All good" : (val < 2 && val >=1) ? "Warning": val === 2 ? "Critical" : "";
},
//color: 'red',
stepSize: 1,
padding: 8
}
}
},
layout: {
padding: {
left: 20,
right: 20,
top: 25,
bottom: 0
}
},
};

var employeeDetailsCtx = document.getElementById("employee-details").getContext("2d");
var employee_details_data = {
labels: ["Late Present", "On Leave", "Training", "Tour"],
datasets: [{
label: "Officer",
backgroundColor: "#5A8DEE",
data: [
...
]
}, {
label: "Staff",
backgroundColor: "#4BC0C0",
data: [
...
]
}]
};
var myoption = {
tooltips: {
enabled: true
},
hover: {
animationDuration: 1
},
onClick: function (evt, i) {
var activePoint = employeeDetailsBarChart.getElementAtEvent(evt)[0];
var data = activePoint._chart.data;
var datasetIndex = activePoint._datasetIndex;
var label = data.datasets[datasetIndex].label;
var value = data.datasets[datasetIndex].data[activePoint._index];
e = i[0];
var x_value = this.data.labels[e._index];
console.log(x_value)
console.log(label)
console.log(value)
},
animation: {
duration: 1,
onComplete: function () {
var chartInstance = this.chart,
ctx = chartInstance.ctx;
ctx.textAlign = 'center';
ctx.fillStyle = "rgba(0, 0, 0, 1)";
ctx.textBaseline = 'bottom';
this.data.datasets.forEach(function (dataset, i) {
var meta = chartInstance.controller.getDatasetMeta(i);
meta.data.forEach(function (bar, index) {
var data = dataset.data[index];
ctx.fillText(data, bar._model.x, bar._model.y - 5);
});
});
}
}
};
var employeeDetailsBarChart = new Chart(employeeDetailsCtx, {
type: 'bar',
data: employee_details_data,
options: myoption
});