Create Chart Axes with D3 v4

Share this video with your friends

Social Share Links

Send Tweet
Published 8 years ago
Updated 5 years ago

Most charts aren’t complete without axes to provide context and labeling for the graphical elements being displayed. This lesson introduces D3’s APIs for creating, customizing, and displaying axes while building on topics from previous lessons.

[00:00] Now that we've made room for our axes, let's see how we actually create them and add them to our chart.

[00:05] The first thing we're going to do is come down here. We need to create a scale. We'll start with a linear scale since that's the most straightforward type. We're just going to give it a domain of 0to 100 and a range of height to 0The reason we're doing that in reverse is because the Y dimension runs from top to bottom in SVG. So we want our minimum value at the bottom of the chart and our maximum value at the top.

[00:31] Now that we've created the scale, we can create the axis itself. I'm going to say var Y axis equals D3.axisleft and then pass in our scale. Now we've created the scale and we've created the axis, but we still need to add the axis to the chart.

[00:47] The way we're going to do that is using the call method that we've seen previously. We're just going to add this directly to the SVG by saying SVG.call and then passing in the axis instance. There we can see our axis actually shows up.

[01:01] It's a little cramped, so let's give a little bit more margin here. We'll give a top margin so we don't get cut off in that direction. Then we'll double the left margin, so we've got plenty of room here.

[01:12] Now we can see that 100 is at the top of our chart, 0is at the bottom just like we expected. These labels are generated based on your domain. If we change that to 500, you can see that we do get that updated. Now we've got to tick every 50 units. You can imagine what happens if you go to 5,000. It's going to update things on its own.

[01:37] What D3 does is it looks at the domain and it creates what it thinks is a reasonable number of ticks for the area and the domain. If you have needs different from what it creates by default though, there are a couple of ways that you can specify how you want it to behave differently.

[01:54] If we were to say .ticks here at the end of our call when we create the axis and passed in five, it'll change how many ticks it creates. You'll notice that there are actually six ticks here if you're counting the minimum and the maximum. That's because ticks is just a suggestion. D3 is still going to use its best judgment on how to place and format these ticks.

[02:21] You can also provide an optional formatting string for the ticks. There's a ton of different options, and it can get really complex. We'll look at a couple just to see basically how it works.

[02:33] If you pass a percent sign, you can see that these now get presented as percentages, but they're crazy big percentages. That's because if you have percentage formatting, it's expected that your domain has been normalized to run from zero to one. If we save that, now we get 100 as our maximum down to 0as our minimum.

[02:54] We can also do things like specifying how many decimals we want. If we keep our domain as zero to one here and specify that we always want decimals, we can change that from two to one and just get one decimal. You really have a lot of control using these different formatting values. Like I said, there's a ton of options. You'll need to do a little bit of looking through the documentation on exactly how to do what you want.

[03:19] The last thing I want to show is that if you have a really large domain so that 1E6 is one million here. Our domain is running from zero to one million. If we change this to an S, we'll actually get these big M notation where we've got .6 M and that sort of thing. If we get rid of that formatting, it's just going to show those raw values. Let's make room for it here. Then you've got 200,000, 400,000, that kind of thing. Let's switch this back. We'll go back to 100 here. There we go.

[03:58] The last thing I want to show regarding this is that if you do need to actually control the exact items that you have ticks for, you can specify that by calling tick values. In this case, instead of suggesting to D3 what you want to be shown on the axis, you can give it the exact values. Maybe we want 8, 19, 43, and 77 on our axis. There we get those exact values.

[04:26] I don't know why you would choose those values, but sometimes you want that finer grain of control. Tick values is what's going to provide it.

[04:35] That's our Y axis again using a linear scale. But one of the other places D3's axis methods are really handy is when you're dealing with dates and times. I'm going to put this back to the default and save that. Then we're going to create a time scale.

[04:51] We're going to come down here. We're going to say var, X scale equals D3.scaletime. Since this is a time scale, our domain needs to hold date objects. For now we'll just do January 1st to February 1st of this year. Then our range is going to be zero to our width.

[05:10] Now that we have our scale, we can create our axis. We're going to say var X axis equals D3.axisbottom this time and pass in our X scale. This scale is going to be oriented so that the labels are on the bottom.

[05:26] Now we can say SVG.call X axis, save that, and things are a little screwy. In fact, you can see that our labels are in fact on the bottom here, but it's at the top of our chart. To fix this, we actually need to use another graphics container and move this down to the bottom. We're going to say SVG.appendG. Then we'll transform that down to the bottom using an ES6 template stream.

[05:56] Now things are in the right position, but obviously it's a little crowded here. Let's use our ticks again to say that we just need about five labels. There we go. That looks better. We'll give ourselves a little more room here to avoid that cutoff. Now you can see that we have date labels here that are more or less evenly spaced out. We have January 3rd, 10th, 17th, 24th, and 31st. Every seven days, we've got a tick and a label.

[06:26] If we were to adjust this so that our domain actually only runs from January 1st at midnight to January 2nd, you can see that it again updates intelligently. It uses 2016 for the beginning of the year there, and then we have 6:00 AM, 12:00 PM, and so on. It will again intelligently update based on these date intervals.

[06:50] Let's show a domain even smaller, January 1st 6:00 AM to January 1st at 12:00 PM. There we've got intervals of one hour. When you're dealing with time scales, you can pass in these different time intervals into ticks instead of just a number. We can say D3.timeminute.every45.

[07:22] That's a little crowded. Let's make this a shorter range. Let's make it a three hour window here, from 6:00 AM to 9:00 AM. We want to tick every 45 minutes. In fact we get 6:00 AM, 6:45 AM, 7:00 AM, 7:45 and so on. There's still a little bit of D3's own choices in there, the fact that it's doing every hour on the hour, but we do have a tick for every 45 minutes after the hour.

[07:55] You can probably imagine that there's nearly infinite ways to customize this and tweak it, but that's all we'll look at now in terms of creating the labels.

[08:04] I'm going to go ahead and turn this back to just a simple interval and look at a couple last things for how to alter the presentation of these axes. I'm going to increase our bottom margin here a bit to give us a little more room to work with.

[08:21] The first thing I want to show is, for our bottom axis here, we've got our ticks specified. We can actually also specify the tick size. If we say tick size is 20, it's going to give us longer ticks here so that we can actually control that.

[08:38] If you want even finer grain control, you can separately specify the inner and outer tick sizes. Maybe we want our inner ticks to be 10 and our outer ticks to be 20. Now we see...Well, it's sort of hard to see there.

[08:54] To show that it actually does work, we can do tick padding which is another option that we have here and set that to 15. You can see the tick padding is based on the smallest tick size. Everything is now aligned here on labels, but we've got these longer ticks on the outside.

[09:14] Those are our different ways of modifying how the ticks and labels are arranged on the chart...

Sergey
Sergey
~ 6 years ago

Hello there. Question about .ticks(d3.timeMinute.every(45)). Why aren't intervals between ticks the same? I'd expect ticks to be: 6am 6:45 7:30 8:15 ... But instead it shows: 6am 6:45 7am 7:45 8am 8:45 ... Thanks in advance!

Sergey
Sergey
~ 6 years ago
Markdown supported.
Become a member to join the discussionEnroll Today