D3 allows you to bind arbitrary data to a Document Object Model (DOM), and then apply data-driven transformations to the document.
Let's select all divs on the page and bind an array of integers to it.
If we hard-code four divs on that page we can use the values specified in
data() to set attributes of selected divs.
We will be using the
text() method which sets the content of a div.
The value that is bound to current element will be passed as a first argument of the function that's passed to
text(). If you are extremely confused at this point just look at this example and try not to think about the space badgers.
In a similar way we can modify other attributes of selected elements.
Usually, you will want to select something more specific. For example all divs within some element. Selectors work in a way you'd expect them to.
We could be using SVG element instead of divs - there's really no difference. But because time is of the essence let's just stick with divs and use some CSS to make it look half-decent. We're using ES6 we can use the arrow functions to tidy up the code.
Much better! Let's make it a bit more dynamic, it will help us understand how things work.
Instead of having a hardcoded value for
dataset let's pull it out from an input field.
For the sake of simplicity we can just use
split() to convert it from a string to an array of values.
We also need to wrap our whole code in a function and then set it as
onClick action for the update button.
Try clicking it and changing the values in the input box and see what happens.
You will notice that there are always exactly four bars, even before we click the update button. This is because we hard-coded them into our HTML code during the first step.
Until now we only used the default (a.k.a. update) selection, which contains the elements for which we have both data and matching DOM elements.
To add new bars we will use the
enter() selection. This selection contains 'ghost' elements for each dataset value that does not have a matching element. (This is not a completely correct explanation, but it's good enough for now.). We can then use
append() to add the missing elements.
This also means that we don't have to hard-code any divs into the DOM, so the root element of our chart can now be just
We still need to remove the extra elements if the dataset is smaller than the number of DOM elements. This is where the
exit() selection kicks in. In this case we don't have to change any attributes so a simple call to
remove() will be good enough.
And that's it, you are saved from the badgers and can return home. Here is the final result: