# Line Chart in Flutter – WeightTracker 6

In this post I will go through the process of creating line chart for my WeightTracker app. Since there is no official support for drawing charts yet, we will do it on our own . Expected result is a Widget displaying history of weight entries from last month. X axis will represent time (days) while Y axis will represent values (weights). Let’s get to it!

## Introduction

At first, I created *StatelessWidget* which contains *CustomPaint*. This Widget can be used anywhere, in my case, I will display it inside a Card. I won’t discuss logic connected to preparing data since it is strictly connected to my business logic (see the full code if interested).

Next step is to create *ChartPainter *– a class that inherits from *CustomPainter* and is our most important class since there all painting will take place:

I created 4 fields (*leftOffsetStart, topOffsetEnd, drawingWidth, drawingHeight*) to calculate the area of drawings. We don’t want to use the full size we get because we need space for labels and small margins.*ShouldRepaint* method ideally would check if we really should repaint our canvas but this is not important in the scope of this post.

. . .

## Calculating Y range

Since we are not using any charting library, we have to calculate the range of values we would like to present. I assumed, that my chart will have 5 horizontal lines each representing integer value.

Since there is no *Pair* in Dart and I wanted to return min and max value with one method, I had to use *Tuple2* from Tuple package. To do that I needed to include Tuple package in *pubspec.yaml* file:

When it comes to calculating border values, at first I get ceil from max weight as max value. After that I calculate difference between min and max value. Then I need to change minimal value so that difference between min and max values is multiple of 4 (number of horizontal values – 1). Let’s crack that by example:

Let’s say that minimal weight in our data equals 7.4 and maximal 13.8.**maxLineValue** = (13.8).ceil() = **14**

difference = 14 – (7.4).floor() = 14 – 7 = 7

toSubtract = 4 – (difference % 4) = 4 – (7%4) = 4 – 3 = 1**minLineValue** = (7.4).floor() – toSubtract = 7 – 1 = **6**

Now having 5 lines, their labels would be accordingly: **6**, 8, 10, 12, **14**.

. . .

## Drawing horizontal lines and labels

Let’s get to drawing horizontal lines and labels. We’ll create grey *Paint* for lines and inside a *for *loop we will draw them on *canvas:*

At first, we need to calculate the weight step between every horizontal line. According to previous example it would be equal to 2:

Then we calculate the pixel difference between lines based on drawing height and number of lines. We will use that value so that every line will have Y-offset equal to numberOfLine*offsetStep :

When we get those, we can paint a line. To do that we need to call *Canvas’ **drawLine* method with *Offsets* of starting and ending points. The ‘+5’ is there because I wanted to move chart a bit down:

The last thing we need to do is add labels to horizontal lines. Most of code are just aesthetic patches:

Right now this is what we got:

. . .

## Drawing bottom labels

Process of creating bottom labels is pretty similar to left-side labels.

Instead of iterating through the number of lines, we do it by weeks (7 days), starting from today and moving 7 days backward with each iteration. To calculate left offset we simply divide drawing width by the number of days in total and multiplying it by number of days in current iteration:

. . .

## Drawing actual data lines

At first we are creating blue *Paint *for our lines. Then for every entry we are going to draw line connecting that entry to the next one and we will draw circle (a dot) representing next entry. At last we will add bigger dot to the first entry, this dot will represent current state (since in my example first value is the most recent one and last value is the oldest):

To draw those lines, we need offsets representing values in our data. The first thing to do is getting relative x and y positions which tell us in what ‘part’ of drawing area point will be placed (from 0 to 1 where 0 is the beginning and 1 is the end). To calculate actual offset we need to get a starting position (our margin) and multiply the relative value with actual drawing size. As a result, we get *Offset* representing one entry:

. . .

## Wrapping up

Having everything together, we end up with chart that looks like this:

And that’s it!

Even if the full process may look complicated, once you get into it, it gets pretty easy, so until we have an official charting library we can still do charts on our own.

I didn’t mention every aspect of my implementation (mostly connected to data transformation). If you are interested in more details, see my GitHub repository for more.

Cheers! ?

#### Marcin Szałek

Founder of Fidev

Flutter enthusiast since Alpha release in 2017. Believes that sharing is caring, which lead him to start a technical blog dedicated fo Flutter in its early days. Loves to see beautiful designs become real apps and is willing to help make it happen. Enjoys sunny beaches far from home.

## Join the newsletter!

Join the newsletter to keep track with latest posts and get my special Widget to animate views' entrances without any hassle for FREE!

## Check out other posts!

## Stripe Checkout in Flutter Web

Flutter Web is getting more mature every day. If you want to accept payments using Stripe Checkout in your Flutter web application, this article is just for you!

## Stripe Checkout in mobile Flutter app

Have you ever struggled to integrate card payments into your mobile Flutter app? If so, today is your lucky day! In this post, I present how to use Stripe Checkout in the Flutter app without any hassle!

## Interacting with Widgets using Framy

Have you ever developed a widget or a page and you wanted to make sure it works correctly in different scenarios but then it turned out that you can’t just reproduce all the cases you want to cover? Framy may solve such problems!

good job, it’s better

Hi:

Great post !!

How could I make the chart to be updated and smoothly redraw with a new set of data ? Say like the new price of a stock * .

Thanks

Hey, I’m glad you like it 🙂

When it comes to “smooth redrawing”, the only idea I have is adding an animation, and then animating every node from 0 to 100% of its value (so it would just naturally rise from X-axis. I think adding animations to just the new points would be much more advanced.

Thanks !

hey can i change colour of line charts when it extends to a limi t t?????plzzhelp

Well, show me a link to what you’ve tried and maybe we can find something. 🙂

Hi Marc,

Thanks for the reference!

I have an app that logs data for every hour, then I need to sum the most recent data value with the previous ones if they are at the same date time before showing them to your chart reference,

any idea how to solve the problem with this line chart reference?