Side by Side Graph D3.js - d3.js

I'm new to D3 so please be patient.
I am trying to build a side by side graph in d3.js. I'm a little stuck trying to add 'rect' (rectangle svg) inside of 'g' (groups). I'm not sure if the json data is formated incorrectly. Or if I am missing some fundamental step.
Json object = data;
var data = [
{
dontknow: 3,
neutral: 0,
optimistic: 62,
pessimistic: 15,
veryoptimistic: 14,
verypessimistic: 6,
year: 2013
},
{
dontknow: 10,
neutral: 4,
optimistic: 42,
pessimistic: 55,
veryoptimistic: 64,
verypessimistic: 5,
year: 2013
}
]
And my code looks like this:
var svg = d3.select("#graph5")
.append("svg")
.attr("width", w)
.attr("height", h);
//Create groups
var g = svg.selectAll('g')
.data(data)
.enter()
.append('g');
//Add rectangles to each g
var rect = g.selectAll("rect")
.data(function(d){ return d })
.enter()
.append("rect")
.attr('width', function(d){ d.year });
The SVG and 'g' elements appear to be created correctly and correct data is associated with the 'g' but the final rect call doesnt seem to work.
Any ideas?
Thanks

Looks like the object keys were causing an issue.
By utilising the inbuild 'd3.values(d)' function I was able to create the rects.
Here is the code (Notice the second line difference):
var rect = g.selectAll("rect")
.data(function(d){ return d3.values(d) })
.enter()
.append("rect")
.attr('width', function(d,i){ console.log(d); return d });

Related

How add legend to a bar-chart based on data in an object dynamically (d3)

I am really new to the realm of D3 and based on the book of 'Interactive Data visualization for the web', I managed to create a Bar chart which is mostly based on the code from the following link.
The problem is I don't manage to add a legend to my bar chart based on an object dynamically.
I have tried consulting youtube videos and other stackoverflow questions related to 'adding a legend to a bar-chart', however in my opinion I couldn't find the question concerning how one is able to retrieve keys from an array of objects and use the data to add as an legend to the bar-chart.
For now all my bars also have the same color, see the second code below.
See the code below for the formatting of my object which is embedded in an array.
The name 'key' and 'value' are fixed, while the amount of the objects and their corresponding name and value differ after an click event of the user ( which determines which variables will be included in the object).
The following example is able create a legend, however in this case the formatting of the object is somehow different than in my case and my current knowledge of D3 is limitd, so I have no idea in which ways I have to adapt the code.
2: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenTotaal", value: 490}
3: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_M_nZakelijkeDienstverlening", value: 165}
4: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_R_uCultuur_Recreatie_OverigeDiensten", value: 120}
5: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_K_lFinancieleDiensten_OnroerendGoed", value: 15}
6: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_ALandbouw_BosbouwEnVisserij", value: 0}
7: {key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_H_p_JVervoer_InformatieEnCommunicatie", value: 85}];
Based on the code from the book and accounting for other variables, I have currently the following code for visualizing a bar chart, in which the values (see object above) are shown in the bar charts and the color of the bar are all blueish. However there is not yet an legend included in my current code. Therefore I am wondering how one is able to dynamically create a legend based on the 'keys' ( in my case)in the object and represent the corresponding color bound to the bars. I would like to achieve the lowest image which I have drawn a sketch of.
var svg = d3.select("#barchart")
.select("svg")
.remove("svg");
//Width and height
var w = 600;
var h = 250;
var padding=20;
var xScale = d3.scaleBand()
.domain(d3.range(dataset.length))
.rangeRound([w - padding,padding ])
.paddingInner(0.05);
var yScale = d3.scaleLinear()
.domain([0, d3.max(dataset, function (d) {
return d.value;
})])
.range([padding,h - padding]);
console.log("yscale",yScale);
//Define key function, to be used when binding data
var key = function (d) {
console.log("key", d);
return d.key;
};
// d3.select("svg").remove();
//Create SVG element
var svg = d3.select("#barchart")
.append("svg")
.attr("width", w)
.attr("height", h);
console.log("svg", svg);
//Create bars
svg.selectAll("rect")
.data(dataset, key) //Bind data with custom key function
.enter()
.append("rect")
.attr("x", function (d, i) {
return xScale(i);
})
.attr("y", function (d) {
return h - yScale(d.value);
})
.attr("width", xScale.bandwidth())
.attr("height", function (d) {
return yScale(d.value);
})
// .attr("data-legend", function (d) { return d.key })
.attr("fill", function (d) {
return "rgb(0, 0, " + (d.value * 10) + ")";
});
//Create labels
svg.selectAll("text")
.data(dataset, key) //Bind data with custom key function
.enter()
.append("text")
.text(function (d) {
return d.value;
})
.attr("text-anchor", "middle")
.attr("x", function (d, i) {
return xScale(i) + xScale.bandwidth() / 2;
})
.attr("y", function (d) {
return h - yScale(d.value) + 14;
})
.attr("font-family", "sans-serif")
.attr("font-size", "11px")
.attr("fill", "white");
If I understood correctly this is what you should need.
Plunker with working code.
First of all I would encourage to use an margin object which will allow better flexibility when dealing with charts
var margin = {
top: 20,
right: 20,
bottom: 20,
left: 20
};
We want to display the data with an odinal scale from the data and example you provided.
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenTotaal", value: 490}
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_M_nZakelijkeDienstverlening", value: 165}
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_R_uCultuur_Recreatie_OverigeDiensten", value: 120}
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_K_lFinancieleDiensten_OnroerendGoed", value: 15}
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_ALandbouw_BosbouwEnVisserij", value: 0}
{key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_H_p_JVervoer_InformatieEnCommunicatie", value: 85}];
Taking into account that probably the first element is a sum of sorts of the dataset I think it shouldn't be included in the chart since it is an aggregation of the elements we want to display.
(In case you need to display it as an element you should be able to do it quickly after reviewing the answer)
The element structure in your dataset is the following:
{
key: "bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_H_p_JVervoer_InformatieEnCommunicatie",
value: 85
}
The domain of our xScale should be all the key values in our dataset, since the key is a huge string, I created a custom property in each element called label
{
key:
"bedrijfsvestigingen_Sbi2008_BedrijfsvestigingenNaarActiviteit_M_nZakelijkeDienstverlening",
label: "Business Services",
value: 165
}
Lets create our scale with the correct domain and range:
var xScale = d3
.scaleBand()
.domain(dataset.map(d => d.label)) // All our label properties
.rangeRound([0, w - margin.left - margin.right]) // This scale will map our values from [0, width - margin.left - margin.right]
.paddingInner(0.05);
The yScale was almost correct, we just need to change it a little to use our margin object and use the correct range
The range must start from 0, if we used padding as the starting point our values will have an offset, since our values would be mapped from [padding, h - padding]. If we wanted to display a zero the value would be mapped to the padding value, if this is way you want to show the information keep it that way. In this case we will modify the scale.
var yScale = d3
.scaleLinear()
.domain([
0,
d3.max(dataset, function(d) {
return d.value;
})
])
.range([0, h - margin.top - margin.bottom]);
Next we will create a function to get the desired value from our elements
var xKey = function(d) {
return d.label;
};
Add our svg with some visual cues to help visualizing the way the elements are layed out:
var svg = d3
.select("#barchart")
.append("svg")
.style("background", "rgb(243, 243, 243)")
.style("border", "1px dashed #b4b4b4")
.attr("width", w)
.attr("height", h);
We want to use a margin, so lets use a group tag to achieve this, we could individually set the margin in each group/element we desired but I find this way simpler and clearer
var g = svg
.append("g")
.attr("transform", `translate(${margin.left}, ${margin.top})`);
We will need the width and height of the chart with the margins taken into account, lets define them really quick:
const customWidth = w - margin.left - margin.right;
const customHeight = h - margin.top - margin.bottom;
Let us add a rect to show where will our rects will be displayed:
g.append("rect")
.attr("fill", "#e3e3e3")
.attr("width", customWidth)
.attr("height", customHeight);
Lets deal with the rect creation, in your code you had a custom fill function which modified the b value within the RGB color values. In this case since we are dealing with categorical data we will use an array of colors for the rects.
g.append("g")
.attr("class", "rect__container")
.selectAll("rect")
.data(dataset, xKey) //Bind data with custom key function
.enter()
.append("rect")
.attr("x", function(d, i) {
return xScale(xKey(d)); // use our key function
})
.attr("y", function(d) {
return customHeight - yScale(d.value); // use our custom size values
})
.attr("width", xScale.bandwidth())
.attr("height", function(d) {
return yScale(d.value);
})
.attr("fill", function(d, i) {
return d3.schemeCategory10[i]; // use an array of colors and use the index to decide which color to use
});
We have two options to show the labels of the chart:
We can create an x-axis or the desired legends. We will do both since it won't affect the outcome of the chart and either one of them can be removed.
var margin = {
top: 20,
right: 300, // modifiy our margin to have space to display the legends
bottom: 50,
left: 20
};
var legendElement = g
.append("g")
.attr("class", "legend__container")
.attr("transform", `translate(${customWidth}, ${margin.top})`) // set our group position to the end of the chart
.selectAll("g.legend__element")
.data(xScale.domain()) // use the scale domain as data
.enter()
.append("g")
.attr("transform", function(d, i) {
return `translate(${10}, ${i * 30})`; // provide an offset for each element found in the domain
});
legendElement
.append("text")
.attr("x", 30)
.attr("font-size", "14px")
.text(d => d);
legendElement
.append("rect")
.attr("x", 0)
.attr("y", -15)
.attr("width", 20)
.attr("height", 20)
.attr("fill", function(d, i) {
return d3.schemeCategory10[i]; // use the same category color that we previously used in rects
});
Now lets use the axis approach:
// create axis
var x_axis = d3.axisBottom().scale(xScale);
//Append group and insert axis
g.append("g")
.attr("transform", `translate(${0}, ${customHeight})`)
.call(x_axis);
g.append("g")
.attr("transform", `translate(${customWidth / 2}, ${customHeight + 40})`)
.append("text")
.text("Activities")
.attr("font-family", "sans-serif")
.attr("font-size", "14px")
.attr("font-weight", "bold")
.style("text-transform", "uppercase")
.attr("text-anchor", "middle");
And finally create the labels for the value in our data:
//Create labels
g.append("g")
.attr("class", "text__container")
.selectAll("text")
.data(dataset, xKey) //Bind data with custom key function
.enter()
.append("text")
.text(function(d) {
return d.value;
})
.attr("text-anchor", "middle")
.attr("x", function(d, i) {
return xScale(xKey(d)) + xScale.bandwidth() / 2;
})
.attr("y", function(d) {
return customHeight - yScale(d.value) + 14;
})
.attr("font-family", "sans-serif")
.attr("font-size", "11px")
.attr("fill", "white");

D3 - Using JSON data to create multiple bars

I'm learning D3 and have JSON data. I want to build multiple bars from this JSON data to draw graph like this already built in excel. I can draw one line of Pax_Rev on SVG but I'm not sure how to add other lines from the data. When I do console.log(dataset.length), it shows me 0 which means only one item in dataset which is expected.
<script>
var dataset = [{"Pax_Rev": 1000, "Crg_Rev": 500,
"Fixed_Costs": 800, "Variable_Costs": 200}];
var width = 500;
var height = 1000;
var barPadding = 1;
var svg = d3.select("body")
.append("svg")
.append("g")
.attr("width", width)
.attr("height", height)
.attr("class", "svg")
svg3.selectAll("rect")
.data(dataset)
.enter()
.append("rect")
.attr("x", 0)
.attr("y", function(d){
return height - d.Pax_Rev // How to add other items like Crg_Rev etc?
})
.attr("width", 20)
.attr("height", function(d){
return d.Pax_Rev
});
</script>
As I explained in your previous question, this is the expected behaviour. Since you have just one object in your data array, D3 "enter" selection will create just one element.
If you look at the API, you'll see that selection.data():
Joins the specified array of data with the selected elements[...] The specified data is an array of arbitrary values (e.g., numbers or objects). (emphases mine)
Therefore, we have to convert that huge object in several objects. This is one of several possible approaches:
var dataset = [{
"Pax_Rev": 1000,
"Crg_Rev": 500,
"Fixed_Costs": 800,
"Variable_Costs": 200
}];
var data = [];
for (var key in dataset[0]) {
data.push({
category: key,
value: dataset[0][key]
})
}
console.log(data)
Now, we have a data array, with several objects, one for each bar, and we can create our bar chart.
Here is a demo:
var dataset = [{
"Pax_Rev": 1000,
"Crg_Rev": 500,
"Fixed_Costs": 800,
"Variable_Costs": 200
}];
var data = [];
for (var key in dataset[0]) {
data.push({
category: key,
value: dataset[0][key]
})
}
var svg = d3.select("svg");
var yScale = d3.scaleLinear()
.domain([0, d3.max(data, function(d) {
return d.value
})])
.range([120, 10]);
var xScale = d3.scaleBand()
.domain(data.map(function(d) {
return d.category
}))
.range([40, 280])
.padding(0.2);
var rects = svg.selectAll(null)
.data(data)
.enter()
.append("rect")
.attr("fill", "steelblue")
.attr("x", function(d) {
return xScale(d.category)
})
.attr("width", xScale.bandwidth())
.attr("y", function(d) {
return yScale(d.value)
})
.attr("height", function(d) {
return 120 - yScale(d.value)
});
var yAxis = d3.axisLeft(yScale);
var xAxis = d3.axisBottom(xScale);
svg.append("g").attr("transform", "translate(40,0)").call(yAxis);
svg.append("g").attr("transform", "translate(0,120)").call(xAxis);
<script src="https://d3js.org/d3.v4.min.js"></script>
<svg></svg>

Conditionally fill/color of voronoi segments

I'm trying to conditionally color these voronoi segments based on the 'd.lon' value. If it's positive, I want it to be green, if it's negative I want it to be red. However at the moment it's returning every segment as green.
Even if I swap my < operand to >, it still returns green.
Live example here: https://allaffects.com/world/
Thank you :)
JS
// Stating variables
var margin = {top: 20, right: 40, bottom: 30, left: 45},
width = parseInt(window.innerWidth) - margin.left - margin.right;
height = (width * .5) - 10;
var projection = d3.geo.mercator()
.center([0, 5 ])
.scale(200)
.rotate([0,0]);
var svg = d3.select("body").append("svg")
.attr("width", width)
.attr("height", height);
var path = d3.geo.path()
.projection(projection);
var voronoi = d3.geom.voronoi()
.x(function(d) { return d.x; })
.y(function(d) { return d.y; })
.clipExtent([[0, 0], [width, height]]);
var g = svg.append("g");
// Map data
d3.json("/world-110m2.json", function(error, topology) {
// Cities data
d3.csv("/cities.csv", function(error, data) {
g.selectAll("circle")
.data(data)
.enter()
.append("a")
.attr("xlink:href", function(d) {
return "https://www.google.com/search?q="+d.city;}
)
.append("circle")
.attr("cx", function(d) {
return projection([d.lon, d.lat])[0];
})
.attr("cy", function(d) {
return projection([d.lon, d.lat])[1];
})
.attr("r", 5)
.style("fill", "red");
});
g.selectAll("path")
.data(topojson.object(topology, topology.objects.countries)
.geometries)
.enter()
.append("path")
.attr("d", path)
});
var voronoi = d3.geom.voronoi()
.clipExtent([[0, 0], [width, height]]);
d3.csv("/cities.csv", function(d) {
return [projection([+d.lon, +d.lat])[0], projection([+d.lon, +d.lat]) [1]];
}, function(error, rows) {
vertices = rows;
console.log(vertices);
drawV(vertices);
}
);
function polygon(d) {
return "M" + d.join("L") + "Z";
}
function drawV(d) {
svg.append("g")
.selectAll("path")
.data(voronoi(d), polygon)
.enter().append("path")
.attr("class", "test")
.attr("d", polygon)
// This is the line I'm trying to get to conditionally fill the segment.
.style("fill", function(d) { return (d.lon < 0 ? "red" : "green" );} )
.style('opacity', .7)
.style('stroke', "pink")
.style("stroke-width", 3);
}
JS EDIT
d3.csv("/static/cities.csv", function(data) {
var rows = [];
data.forEach(function(d){
//Added third item into my array to test against for color
rows.push([projection([+d.lon, +d.lat])[0], projection([+d.lon, +d.lat]) [1], [+d.lon]])
});
console.log(rows); // data for polygons and lon value
console.log(data); // data containing raw csv info (both successfully log)
svg.append("g")
.selectAll("path")
.data(voronoi(rows), polygon)
.enter().append("path")
.attr("d", polygon)
//Trying to access the third item in array for each polygon which contains the lon value to test
.style("fill", function(data) { return (rows[2] < 0 ? "red" : "green" );} )
.style('opacity', .7)
.style('stroke', "pink")
.style("stroke-width", 3)
});
This is what's happening: your row function is modifying the objects of rows array. At the time you get to the function for filling the polygons there is no d.lon anymore, and since d.lon is undefined the ternary operator is evaluated to false, which gives you "green".
Check this:
var d = {};
console.log(d.lon < 0 ? "red" : "green");
Which also explains what you said:
Even if I swap my < operand to >, it still returns green.
Because d.lon is undefined, it doesn't matter what operator you use.
That being said, you have to keep your original rows structure, with the lon property in the objects.
A solution is getting rid of the row function...
d3.csv("cities.csv", function(data){
//the rest of the code
})
... and creating your rows array inside the callback:
var rows = [];
data.forEach(function(d){
rows.push([projection([+d.lon, +d.lat])[0], projection([+d.lon, +d.lat]) [1]])
});
Now you have two arrays: rows, which you can use to create the polygons just as you're using now, and data, which contains the lon values.
Alternatively, you can keep everything in just one array (just changing your row function), which is the best solution because it would make easier to get the d.lon values inside the enter selection for the polygons. However, it's hard providing a working answer without testing it with your actual code (it normally ends up with the OP saying "it's not working!").

Adding text on each bar of histogram

I have problem adding text in my histogram. I can do this in more simple example.
I try to do this:
// try to add bar value
var barnum = g.selectAll('text')
.data(layout)
.enter()
.append("text")
.attr('y',-10)
.attr('x',10)
.attr("text-anchor", "middle")
.style("fill","black")
.text('testtest')
.style("pointer-events", "none")
;
barnum.transition();
I can't see any text in my figure. The code include definition is here:
var dateFormat = d3.time.format("%Y-%m-%d");
var g;
var data;
var margin = {top: 30, right: 30, bottom: 80, left: 80},
width = 500 - margin.left - margin.right,
height = 500 - margin.top - margin.bottom;
var cx = 10;
var numberBins = 5;
var dispatch = d3.dispatch(chart, "hover");
function chart(container) {
g = container;
update();
}
chart.update = update;
function update() {
// create hist layout
var hist = d3.layout.histogram()
.value(function(d) { return d.selectvar })
.range([d3.min(data, function(d){ return d.selectvar }) , d3.max(data, function(d){ return d.selectvar }) ])
.bins(numberBins);
var layout = hist(data);
var maxLength = d3.max(layout, function(d) { return d.length });
var widthScale = d3.scale.linear()
.domain([0, maxLength])
.range([0, width])
var yScale = d3.scale.ordinal()
.domain(d3.range(numberBins))
.rangeBands([height, 0], 0)
var colorScale = d3.scale.category20();
// create svg
var rects = g.selectAll("rect")
.data(layout)
rects.enter().append("rect")
rects .transition()
.duration(500)
.attr({
y: function(d,i) {
return yScale(i)
},
x: 50,
height: yScale.rangeBand(),
width: function(d,i) {
return widthScale(d.length)
},
fill: function(d, i) { return colorScale(i) }
});
rects.exit().transition().remove();
// try to add bar value
var barnum = g.selectAll('text')
.data(layout)
.enter()
.append("text")
.attr('y',-10)
.attr('x',10)
.attr("text-anchor", "middle")
.style("fill","black")
.text('testtest')
.style("pointer-events", "none")
;
barnum.transition();
is there something wrong with my way to create svg element? I found out some successful case use append('g') from the beginning. New to d3.js! thank you.
You're using d3.dispatch, which is documented on a page titled Internals. That doesn't mean you shouldn't use it, but rather, it shouldn't be your first choice.
You're correct that there's "something wrong with my way to create svg element" -- you're not creating one! Try:
var svg = d3.select("body").append("svg");
var g = svg.append("g");
At this point, you need to have a good understanding of the DOM, the SVG standard, CSS selectors, and D3's selection API to make things work. You don't tell D3 to put labels on your bars and that's it. You have to instruct it what elements to create, and where, keeping track of translates and offsets and stuff like that. You're best off copying and studying one of Mike Bostock's many examples.
D3 is not learned quickly. You need to invest time learning it before you can make any chart you like.

Why isn't my updated data re-rendering?

I'm playing around with the "update" pattern in D3.js. I am just creating a simple bar graph that will update the data when you press the "Change" button.
My problem is that when you press the "Change" button, the first three rendered bars do not get re-rendered. I debugged and saw that the data was properly applied (__data__ was correct) but the re-application failed.
Here is my code and a link to it in CodePen:
var myData = [ 100, 200, 300 ];
d3.select('body').append('button').text("Change").on("click", function() {
myData = [200, 400, 600, 700, 800, 900, 1000];
update(myData);
});
var svg = d3.select('body').append('svg')
.attr("class", "chart")
.attr("y", 30);
var update = function(data) {
var bars = svg.selectAll('g')
.data(data);
var groups = bars.enter()
.append("g")
.attr("transform", function(d,i) {return "translate(0," + i*25 + ")"});
groups
.append("rect")
.attr("height", 25)
.attr("fill", "pink")
.attr("stroke", "white");
groups
.append("text")
.attr("x", 10)
.attr("y", 18)
.attr("fill", "red");
bars.selectAll("rect")
.attr("width", String);
bars.selectAll("text")
.text(String);
};
update(myData);
It works if you change the .selectAll() in your update selection handling to .select():
bars.select("rect")
.attr("width", String);
bars.select("text")
.text(String);
By using selectAll(), you're accessing the data that is bound to the elements that you're selecting (i.e. the rectangles and text elements), which was bound when you appended the elements. This data hasn't been updated though as you've only updated it for the containing g elements. Using .select() instead also binds the new data to the child elements.
The general pattern that you're using is a nested selection and can be a bit confusing to start with and lead to unexpected results.

Resources