Your Leaflet map should be coming along nicely after the Chapter 5 overview of interactions. In Chapter 6, we dive deeper into Leaflet's inner workings to introduce you to some advanced web programming concepts. For this chapter, you should copy the Chapter 5's folder and continue working on those files. Chapter 6 includes three lessons and culminates in the completion of your Leaflet map.
- In Lesson 1, we contrast procedural versus object-oriented coding with JavaScript by refactoring some existing, suboptimal code from Chapter 4-5.
- In Lesson 2, we refresh our introduction of Leaflet to look at it as a library of classes and use several of these classes to implement UI controls and a temporal legend.
- In Lesson 3, we introduce scalable vector graphics (SVGs)—a graphics format we rely on with D3 in unit-3 and use SVG to build an attribute legend for our Leaflet map example.
After this chapter, you should be able to:
- Understand how JavaScript pseudo-classes and prototypes are used by Leaflet
- Extend Leaflet objects in your script
- Create your own SVG graphics and integrate them with your Leaflet map
JavaScript employs two different computer programming paradigms: procedural programming and object-oriented programming (OOP). Procedural programming uses individual variables, functions, and data structures to give step-by-step instructions to the processor. In contrast, object-oriented programming uses data and methods contained within complex data structures called objects. JavaScript is primarily a procedural language, but has some characteristics that allow advanced programmers to take an object-oriented approach. In the code for your Leaflet map, you have actually already used a bit of both.
The chapter tutorials use the term function to refer to a set of tasks or routines written into the main.js custom script, and the term method to refer to functions that are part of a library such as jQuery or Leaflet. This is an important semantic distinction. Methods are routines that look like functions but are included as part of an object. Functions belong to the domain of procedural programming: they are a set of instructions that can be used anywhere to redirect the flow of execution, and are not designated as part of a particular code package. On the other hand, code libraries use an object-oriented approach to classes and their methods so that specific child classes inherit (as introduced in Chapter 5), or make use of, broader parent classes, requiring the parent method to be written just once for efficiency.
To examine the difference, let's refactor some of the main.js script from the end of Chapter 5. As you hone your programming skills, you will begin to notice places where your code is more lengthy than it needs to be and could be refactored, or revised for efficiency. To start refactoring, look for code that occurs in a very similar form in multiple places in your script. In procedural programming, this repetition can be reduced by consolidating the duplicate code into its own function that can be called from multiple places to return the desired value.
In the updatePropSymbols() function in main.js of the Leaflet map example code, we recreate our popup content for each feature with script very similar to the initial creation of our popups in the pointToLayer() function (Example 1.1).
//CODE FROM pointToLayer() FUNCTION
//build popup content string
var popupContent = "<p><b>City:</b> " + feature.properties.City + "</p>";
//add formatted attribute to panel content string
var year = attribute.split("_")[1];
popupContent += "<p><b>Population in " + year + ":</b> " + feature.properties[attribute] + " million</p>";
//bind the popup to the circle marker
layer.bindPopup(popupContent, {
offset: new L.Point(0,-options.radius)
});
...
//CODE FROM updatePropSymbols() FUNCTION
//build new popup content string
var popupContent = "<p><b>City:</b> " + props.City + "</p>";
//add formatted attribute to panel content string
var year = attribute.split("_")[1];
popupContent += "<p><b>Population in " + year + ":</b> " + props[attribute] + " million</p>";
//update popup with new content
popup = layer.getPopup();
popup.setContent(popupContent).update();
Let's improve our script by consolidating the redundant code into one procedural function that can be called from both pointToLayer() and updatePropSymbols() (Example 1.2). First, note which variables should be passed into the function as parameters: feature.properties and the selected attribute from the dataset. If you assign a different parameter name for any of these variables, make sure you also change the variable name within the function to reflect the parameter name.
function createPopupContent(properties, attribute){
//add city to popup content string
var popupContent = "<p><b>City:</b> " + properties.City + "</p>";
//add formatted attribute to panel content string
var year = attribute.split("_")[1];
popupContent += "<p><b>Population in " + year + ":</b> " + properties[attribute] + " million</p>";
return popupContent;
};
You then can replace the popup creation code in each of the functions where it appears (Example 1.3).
//Example 1.1 line 2...in pointToLayer()
var popupContent = createPopupContent(feature.properties, attribute);
//bind the popup to the circle marker
layer.bindPopup(popupContent, { offset: new L.Point(0,-options.radius) });
...
//Example 1.1 line 18...in updatePropSymbols()
var popupContent = createPopupContent(props, attribute);
//update popup with new content
popup = layer.getPopup();
popup.setContent(popupContent).update();
Our procedural refactoring reduces the size of our code by a few lines and eliminates unnecessary duplication. The function calls are essentially a stand-in for the duplicate script; you still have an ordered set of step-by-step procedures.
An object-oriented approach instead assigns the variables as properties of an object. These properties then can be inherited by other child objects. This special type of "parent" object—similar to a class in many object-oriented languages—is created by a constructor function. Because this function is analogous to a class, it is convention to capitalize the first letter of a constructor function's name. Example 1.4 shows createPopupContent() as a constructor function (renamed PopupContent()).
//Example 1.2 line 1...PopupContent constructor function
function PopupContent(properties, attribute){
this.properties = properties;
this.attribute = attribute;
this.year = attribute.split("_")[1];
this.population = this.properties[attribute];
this.formatted = "<p><b>City:</b> " + this.properties.City + "</p><p><b>Population in " + this.year + ":</b> " + this.population + " million</p>";
};
...
//Example 1.3 line 1...in pointToLayer()
//create new popup content
var popupContent = new PopupContent(feature.properties, attribute);
//bind the popup to the circle marker
layer.bindPopup(popupContent.formatted, {
offset: new L.Point(0,-options.radius)
});
...
//Example 1.3 line 6...in UpdatePropSymbols()
var popupContent = new PopupContent(props, attribute);
//update popup with new content
popup = layer.getPopup();
popup.setContent(popupContent.formatted).update();
In the example above (Example 1.4), the keyword this refers to the function's prototype property. Think of the prototype as the immediate parent of the objects that are instantiated by the constructor function. Every JavaScript object has a prototype—its parent—which defines its properties and methods. Each object instance is created using the special new keyword to call the constructor, passing each required parameter to the constructor function. When new popupContent objects are instantiated in Example 1.4 on lines 21 and 29, these objects inherit the properties and methods assigned to the prototype using this in the constructor. Lines 3-8 assign properties to the prototype based on the parameters passed into the constructor function.
Inheritance is a powerful feature of object-oriented JavaScript. Once an object is created from a constructor function, it has a set of default properties and methods. However, any of the defaults assigned by the constructor can be changed for individual objects created from it. For instance, imagine that you want to temporarily change the styling of your popups to show just the population and use a larger font. You can replace the formatted property of your popupContent object (Example 1.5).
//create new popup content
var popupContent = new PopupContent(feature.properties, attribute);
//change the formatting
popupContent.formatted = "<h2>" + popupContent.population + " million</h2>";
//add popup to circle marker
layer.bindPopup(popupContent.formatted, {
offset: new L.Point(0,-options.radius)
});
Figure 1.1 shows the result of Example 1.5.
Further, imagine that you want to create two styles of popups, one that maintains the original formatting and another that replaces the formatting, giving the user the option to switch between the two. You can create a second popupContent2 with the first popupContent as its prototype using the Object.create() method (Example 1.6).
//create new popup content...Example 1.4 line 1
var popupContent = new PopupContent(feature.properties, attribute);
//create another popup based on the first
var popupContent2 = Object.create(popupContent);
//change the formatting of popup 2
popupContent2.formatted = "<h2>" + popupContent.population + " million</h2>";
//add popup to circle marker
layer.bindPopup(popupContent2.formatted);
console.log(popupContent.formatted) //original popup content
The console.log() statement shows that the first popupContent object has maintained its original format (Figure 1.2). The second popupContent2 object inherits all of the properties and methods of the first, but then has its format changed on line 8 before binding this content to the circle marker layer.
Prototypes and constructor functions are advanced JavaScript concepts, so we do not expect you to rely on them heavily for your Leaflet map. However, consider how you can make use of prototypes and constructors as you begin to plan your final project.
Hopefully, the previous discussion has made the workings of Leaflet a little clearer to you. Leaflet is basically a giant object (L) that contains an umbrella constructor function, L.Class. Each Leaflet "class", such as L.Map, accesses this constructor function to create its particular kind of object. The lowercase methods used to create these classes, such as L.map(), are embedded in Leaflet as aliases of the object instantiation calls (i.e., new L.Map()). For instance, Example 2.1 shows alias methods for L.Map().
L.map = function (id, options) {
return new L.Map(id, options);
};
Each Leaflet class is created using an extend() method built into the L.Class constructor. The extend method in turn is inherited by the child class, its children, and so on. This means that new Leaflet classes can be created by extending any class, and they inherit the methods and properties of all previous class "generations"—as in the case of the L.LayerGroup, L.FeatureGroup, and L.GeoJSON chain discussed in Chapter 4. It also means that you can create custom Leaflet classes in your own script.
The most common use of extend() in Leaflet is the L.Control class. Examine the Custom Control Example in the Leaflet API documentation to see how L.Control.extend() is used to create custom controls in any of the map's four corners. We immediately can apply this method to make our sequence UI controls more accessible to the user by placing them on the map.
Let's start by adapting the example code in the documentation to create a new SequenceControl class within our createSequenceControls() function in main.js (Example 2.2).
//Create new sequence controls
function createSequenceControls(attributes){
var SequenceControl = L.Control.extend({
options: {
position: 'bottomleft'
},
onAdd: function () {
// create the control container div with a particular class name
var container = L.DomUtil.create('div', 'sequence-control-container');
// ... initialize other DOM elements
return container;
}
});
map.addControl(new SequenceControl()); // add listeners after adding control}
The value of the extend() method—as seen in the example above—is that it takes advantage of an existing constructor function within Leaflet to add properties and methods to the class prototype object the same way they could be added to any other kind of object. The options property and onAdd() method are tacked onto the prototype object for the Control class by extend(), and the revised object becomes the prototype for SequenceControl.
The onAdd() method creates the HTML element and child elements for the Leaflet control, along with HTML attributes and event listeners for the Leaflet control. As the name of the method implies, this script is executed when the control is added to the map. Similarly, we can add an onRemove() method to remove elements and event listeners from the DOM when the control is removed. Since we only are adding the control to our map, we do not need to use onRemove(). However, onAdd() always is required for a new Leaflet control.
Following the documentation example, we create a new HTML <div> element within the onAdd() method using Leaflet's DOM Utility and the L.DomUtil.create() method. We also can create the <div> using JavaScript's native document.createElement() method, but L.DomUtil.create() is slightly more convenient because it automatically adds a class name (as in Example 2.2) and optionally assigns the new element to a parent element. We do not want to use jQuery for this, as we need to define the new Leaflet control without adding it to the DOM immediately.
We can use jQuery for the next step, which is to place our "range-slider" inside of the SequenceControl. To do this, we simply move the line that creates the "range-slider" into the onAdd() method, appending it to the container element using jQuery (Example 2.3).
//Example 2.2 line 8
onAdd: function () {
// create the control container div with a particular class name
var container = L.DomUtil.create('div', 'sequence-control-container');
//create range input element (slider)
$(container).append('<input class="range-slider" type="range">');
return container;
}
Figure 2.1 shows the resulting slider as an extended L.Control in the lower-left corner of your map .
You similarly can move the step buttons into the SequenceControl by placing them in onAdd() (Example 2.4).
//Example 2.3 line 1
onAdd: function () {
// create the control container with a particular class name
var container = L.DomUtil.create('div', 'sequence-control-container');
//create range input element (slider)
$(container).append('<input class="range-slider" type="range">');
//add skip buttons
$(container).append('<button class="step" id="reverse" title="Reverse">Reverse</button>');
$(container).append('<button class="step" id="forward" title="Forward">Forward</button>');
return container;
}
Figure 2.2 shows the resulting step buttons as part of the extended L.Control. Note that the slider and step buttons are now part of the same extended Leaflet control called SequenceControl.
We need to make several additional adjustment from our original code to have the sequence UI controls extending L.Control, rather than sit in a separate <div> element outside of the map. First, if you use the slider, you will notice that dragging the marker also causes the map to move—an undesirable response. Likewise, clicking the step buttons in rapid succession causes the map to zoom in. In both cases, executing of the sequence operator triggers a second operator (pan versus zoom) because the map's default event listeners still are active and SequenceControl is part of the map.
We can deactivate the map's mouse event listeners for the area covered by the SequenceControl control container using L.DomEvent.disableClickPropagation() (Example 2.5).
//Example 2.4 line 1
onAdd: function () {
// create the control container with a particular class name
var container = L.DomUtil.create('div', 'sequence-control-container');
...
//disable any mouse event listeners for the container
L.DomEvent.disableClickPropagation(container);
return container;
}
Second, we need to modify our styles in style.css to better position our slider and buttons in the control container. We also remove the right panel. If any of the Example 2.6 styles are unfamiliar, remember that you can look them up on W3Schools.
#map {
height: 400px;
width: 100%;
display: inline-block;
}
...
.sequence-control-container {
width: 350px;
height: 30px;
background-color: rgba(255,255,255,0.8);
padding: 10px;
line-height: 45px;
text-align: center;
border: solid gray 1px;
}
.range-slider {
width: 250px;
}
.step {
width: 40px;
}
.step img {
width: 100%;
height: auto;
}
Lastly, make sure you add the event listeners for your slider and step buttons after adding the controls; otherwise you are trying to attach listeners with to HTML elements that have not been created yet.
onAdd: function (){
//adding slider and step buttons
.....
}
});
map.addControl(new SequenceControl());
//ATTACH LISTENERS HERE
Figure 2.3 shows our beautiful new UI controls.
The final requirement of the Leaflet map assignment is a temporal legend for the sequence operator and an attribute legend for the proportional symbols. Let's start with the more straightforward temporal legend. We will build the attribute legend using SVG in Lesson 3.
First, create a new extended L.Control control for the temporal legend using a new createLegend() function (Example 2.7). Call the extended control LegendControl and place it in the bottom-right corner of the map. You can change the position of all controls to customize your design after you get their functionality working. Add a call to the new createLegend() function in the AJAX callback.
function createLegend(attributes){
var LegendControl = L.Control.extend({
options: {
position: 'bottomright'
},
onAdd: function () {
// create the control container with a particular class name
var container = L.DomUtil.create('div', 'legend-control-container');
//PUT YOUR SCRIPT TO CREATE THE TEMPORAL LEGEND HERE
return container;
}
});
map.addControl(new LegendControl());
};
At this point, you should have a good idea of how to create the temporal legend based on the attribute names in your GeoJSON dataset. Rather than walking through all the steps here, we will let you adapt script for the popups to populate the temporal legend.
Keep in mind that the temporal legend needs to update with each sequence interaction. Thus, we recommend that you add your temporal legend script to the updatePropSymbols() function so that the legend always matches the proportional symbols.
Also, be sure to add styles for your legend container in style.css, otherwise the container will not show up! The legend should match the look and feel of your map elements, such as the basemap tileset, the popup styling, and the sequence UI controls.
Figure 2.4 shows an example temporal legend. You will have the opportunity to check your code in Lesson 3 if you run into any problems with the temporal legend.
Thematic maps typical require an attribute legend to define the meaning of the map symbols. It is conventional for a proportional symbol map to include a legend of nested circles representing the minimum, maximum, and one or more intermediate values.
You could create your attribute legend using static HTML <div> or <img> elements. However, Leaflet uses SVG graphics to dynamically create its circle markers and other vector linework, so it makes sense to replicate the circle markers using SVG for the legend. Additionally, you will need to understand how SVG works to complete the D3 map in the next unit, so a basic introduction to the SVG standard now will help you in the future.
SVG, or Scamaple Vector Graphics, is the web standard vector graphics format. As with vector geospatial data introduced in Chapter 3, vector graphics use points, lines, and polygon fills to represent image elements. In contrast, raster graphics (such as the png or Portable Network Graphics format), much like raster geospatial data, use a continuous grid of pixels.
SVG uses XML markup, making it relatively human-readable and easy to integrate with HTML. Microsoft initially had its own Vector Markup Language (VML), and Internet Explorer was the last major browser to adopt SVG support. Accordingly, IE 8 and below do not support SVG, the primary reason most boilerplate websites check for this browser. Fortunately, there are increasingly fewer users of these older browser versions, so we will not be concerned with this problem in this workbook.
The SVG standard describes the elements and attributes available as part of an SVG drawing. Every SVG graphic begins with the tag <svg>, within which every other SVG element is nested.
SVG provides a link between vector artwork software such as Adobe Illustrator and web graphics. Illustrator can save graphics as SVG, which then can be added to a webpage. Although SVG circles are quite simple to code from scratch, we start with Illustrator in this lesson to demonstrate how the SVG export works. This method can be applied to more complex graphics as well.
Let's start with a simple 180-pixel diameter circle drawn with the Ellipse Tool in Illustrator (Figure 3.1).
Notice we are using the same colors and fill opacity (80%) as our Leaflet circle markers. Save your circle as an SVG using "File"→"Save As". Choose "SVG" from the dropdown menu, choose a logical place to save the file, and click "Save". Accept the defaults on the next window and click "Save".
Once you have saved the SVG file, you can open it in your text editor to view the SVG content (Example 3.1).
<?xml version="1.0" encoding="utf-8"?>
<!-- Generator: Adobe Illustrator 15.1.0, SVG Export Plug-In . SVG Version: 6.00 Build 0) -->
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
<svg version="1.1" id="Layer_1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" x="0px" y="0px"
width="180px" height="180px" viewBox="0 0 180 180" enable-background="new 0 0 180 180" xml:space="preserve">
<circle fill="#F47821" fill-opacity="0.8" stroke="#000000" stroke-miterlimit="10" cx="90" cy="90.001" r="89.5"/>
</svg>
Compare this code to the circle example on the Mozilla SVG documentation page. The SVG content starts with the opening <svg> tag (line 4), with the rest generated by Illustrator and thus can be deleted.
We can copy and paste this <svg> tag and insert it into the <svg> of our boilerplate index.html file to draw the circle in a webpage (Example 3.2). Here, you could add id and class attributes, allowing you to restyle through CSS and modify through JavaScript code.
<body>
<svg id="Layer_1" width="180px" height="180px">
<circle fill="#F47821" fill-opacity="0.8" stroke="#000000" stroke-miterlimit="10" cx="90" cy="90.001" r="89.5"/>
</svg>
</body>
Figure 3.2 renders the circle in a browser.
In addition to importing SVG code to HTML, you also can export SVG graphics created by your script. Simply use the Inspector on the SVG object, then Right-/Command-Click on its code in the HTML tab and choose "Copy"→"Outer HTML" (Figure 3.3). Then paste the code into a new document in your text editor and save it as a file with a .svg suffix. Your can open your new SVG file in Illustrator or any other graphics software that supports SVG. Notably, many news rooms actually generate much of the map through interactive coding with libraries such as D3, exporting different portions of their dynamic graphics as SVGs for static reproduction in print!
Before moving on, let's briefly look at the code on line 3 of Example 3.2. The SVG <circle> element creates our circle. It has several attributes that define and style it. These are:
-
fill: the fill color of the circle -
fill-opacity: the fill opacity of the circle -
stroke: the stroke color of the circle -
stroke-miterlimit: the limit on the ratio of miter length to stroke width at line segment joins. Irrelevant for circles but added by Illustrator as a generic attribute for strokes. -
cx: the x-axis (horizontal) coordinate of the circle's center point -
cy: the y-axis (vertical) coordinate of the circle's center point -
r: the radius of the circle
You might expect that Leaflet passes the options from the L.CircleMarker options object to SVG <circle> elements that it creates. In fact, this class does create SVG elements, but they are <path> elements rather than <circle> elements. We can see this using the Inspector on our Leaflet map (Figure 3.4).
An SVG <path> element can be used to draw any shape of line or polygon. As you can see, the stroke and fill options are set by the options object specified in the pointToLayer() function, but the circle shape definition is a bit more complicated. Instead of center coordinates and a radius, the circle's shape is set by the d attribute, which takes a coded string of coordinate values (Example 3.3).
d="M341,186.92441514450513A15.07558485549488,15.07558485549488,0,1,1,340.9,186.92441514450513 z"
If you are interested in deconstructing this code, visit the Mozilla documentation page for the d attribute. You will become more familiar with the d attribute in unit-3 when using D3; we stick with the SVG <circle> element for our attribute legend.
In case you are wondering about the <g> elements in Figure 3.4, these are group elements that contain each <path>, but could contain more than one other element and be used to uniformly style all of their child elements.
Returning to our attribute legend, we can use pseudocode to clarify our tasks (Example 3.4).
//PSEUDOCODE FOR ATTRIBUTE LEGEND
Step 1. Add an `<svg>` element to the legend container
Step 2. Add a `<circle>` element for each of three attribute values: min, max, and mean
Step 3. Assign each `<circle>` element a center and radius based on the dataset min, max, and mean values of all attributes
Step 4. Create legend text to label each circle
Step 1 in Example 3.4 dynamically adds an <svg> element to the legend container in our createLegend() function from Lesson 2 (Example 2.7). Since we also dynamically add each circle in Step 2, write out the opening <svg> tag as a string and assign it to the variable svg. For now, specify only the id, width, and height attributes for the <svg> element . Then append svg to the container using jQuery (Example 3.5).
//Example 2.7 line 1...function to create the legend
function createLegend(attributes){
var LegendControl = L.Control.extend({
options: {
position: 'bottomright'
},
onAdd: function () {
// create the control container with a particular class name
var container = L.DomUtil.create('div', 'legend-control-container');
$(container).append('<div class="temporalLegend">Population in <span class="year">1980</span></div>');
//Step 1: start attribute legend svg string
var svg = '<svg id="attribute-legend" width="130px" height="130px">';
//add attribute legend svg to container
$(container).append(svg);
return container;
}
});
map.addControl(new LegendControl());
};
Let's take a look at our legend control via the Inspector (Figure 3.5).
Step 2 of the pseudocode creates each example proportional circle for the legend. Since each legend circle will have its own id attribute to access it for scaling, it makes sense to create an array of circle names over which the loop can iterate (Example 3.6).
//Example 3.5 line 15...Step 1: start attribute legend svg string
var svg = '<svg id="attribute-legend" width="130px" height="130px">';
//array of circle names to base loop on
var circles = ["max", "mean", "min"];
//Step 2: loop to add each circle and text to svg string
for (var i=0; i<circles.length; i++){
//circle string
svg += '<circle class="legend-circle" id="' + circles[i] +
'" fill="#F47821" fill-opacity="0.8" stroke="#000000" cx="65"/>';
};
//close svg string
svg += "</svg>";
//add attribute legend svg to container
$(container).append(svg);
On line 10 of Example 3.6, we assign the circle id attribute based on the current value of the array. Other attributes are similar to the code on line 3 of Example 3.2, except that we have left out the unnecessary stroke-miterlimit attribute as well as the necessary cy and r attributes. We have left out the latter two because these will be assigned dynamically depending on the dataset values for each attribute. Because they are left out, no circles will appear yet in the legend, but if you were to inspect the legend, you could see that the <circle> elements are all present in the DOM.
After the loop, we must add a closing </svg> tag to the string (line 15) before appending it to the control container.
For Step 3 of our pseudocode, we want to size and center our legend's circles based on the overall maximum, mean, and minimum values of our dataset across all years. Since we already are calculating the overall minimum value of our dataset for the Flannery perceptual scaling of our symbols, we can update our calcMinValue() function to calculate the maximum and mean statistics as well. Rename the calcMinValue() function to the broader calcStats() and store the calculated max, mean, and min values as properties in a new, globally accessible dataStats object that we can access from our createLegend() function.
//Line 1: variables we want to access from multiple functions
var map;
var dataStats = {};
.....
function calcStats(data){
//create empty array to store all data values
var allValues = [];
//loop through each city
for(var city of data.features){
//loop through each year
for(var year = 1985; year <= 2015; year+=5){
//get population for current year
var value = city.properties["Pop_"+ String(year)];
//add value to array
allValues.push(value);
}
}
//get min, max, mean stats for our array
dataStats.min = Math.min(...allValues);
dataStats.max = Math.max(...allValues);
//calculate meanValue
var sum = allValues.reduce(function(a, b){return a+b;});
dataStats.mean = sum / allValues.length;
}
.....
function getData(map){
//load the data
$.getJSON("data/MegaCities.geojson", function(response){
.....
//calling our renamed function
calcStats(response);
.....
});
};
Now that we have calculated the statistics we need to size and center our circles, we assign radius and vertical center coordinates to each circle's HTML string within our loop in the createLegend() function (Example 3.8).
//array of circle names to base loop on
var circles = ["max", "mean", "min"];
//Step 2: loop to add each circle and text to svg string
for (var i=0; i<circles.length; i++){
//Step 3: assign the r and cy attributes
var radius = calcPropRadius(dataStats[circles[i]]);
var cy = 130 - radius;
//circle string
svg += '<circle class="legend-circle" id="' + circles[i] + '" r="' + radius + '"cy="' + cy + '" fill="#F47821" fill-opacity="0.8" stroke="#000000" cx="65"/>';
};
//close svg string
svg += "</svg>";
The radius should be calculated the same way as for the symbols on the map, so we use our calcPropRadius() function to get each circle's radius based on its value (line 8).
Since we would like the circles to "nest" with each having a common bottom point, we can set the center y coordinate (cy) by subtracting the radius from the SVG height, minus an extra pixel for the circle stroke (line 10). This makes them appear to "grow up" from the bottom rather than down from the top. With all of our circle attributes set, we can now see our legend circles (Figure 3.6).
The SVG container in Figure 3.6 is oversized given the sizes of our circles. The horizontal position of the circles should be adjusted to make room for the legend text we will create for Step 4 of our pseudo code. For the example legend, we will adjust the following:
-
In
createLegend(), change the<svg>element'swidthto"160px"andheightto"60px" -
In
createLegend(), change each circle'scxattribute value to"30"andcyattribute to"59 - radius"
This gives us a much neater legend that no longer covers some of our map symbols (Figure 3.7). You should adjust these values based on the maximum circle size on your own legend.
Now we are ready for Step 4 of the pseudocode. We can create text within an SVG using <text> elements. Return to the createLegend() function and add to the SVG to create our legend text (Example 3.9).
//Example 3.8 line 4...loop to add each circle and text to SVG string
for (var i=0; i<circles.length; i++){
//Step 3: assign the r and cy attributes
var radius = calcPropRadius(dataStats[circles[i]]);
var cy = 59 - radius;
//circle string
svg += '<circle class="legend-circle" id="' + circles[i] + '" r="' + radius + '"cy="' + cy + '" fill="#F47821" fill-opacity="0.8" stroke="#000000" cx="30"/>';
//evenly space out labels
var textY = i * 20 + 20;
//text string
svg += '<text id="' + circles[i] + '-text" x="65" y="' + textY + '">' + Math.round(dataStats[circles[i]]*100)/100 + " million" + '</text>';
};
//close svg string
svg += "</svg>";
//add attribute legend svg to container
$(container).append(svg);
This adds a <text> element with a unique id and content for each circle to the SVG (line 8). Notice that we evenly space out each <text> element's y coordinate attribute for readability. This completes our legend and the basic requirements for the Leaflet map (Figure 3.8)!
function getCircleValues(attribute) {
//start with min at highest possible and max at lowest possible number
var min = Infinity,
max = -Infinity;
map.eachLayer(function (layer) {
//get the attribute value
if (layer.feature) {
var attributeValue = Number(layer.feature.properties[attribute]);
//test for min
if (attributeValue < min) {
min = attributeValue;
}
//test for max
if (attributeValue > max) {
max = attributeValue;
}
}
});
//set mean
var mean = (max + min) / 2;
//return values as an object
return {
max: max,
mean: mean,
min: min,
};
}
function updateLegend(attribute) {
//create content for legend
var year = attribute.split("_")[1];
var content = "Population in " + year;
//replace legend content
$("#temporal-legend").html(content);
//get the max, mean, and min values as an object
var circleValues = getCircleValues(attribute);
for (var key in circleValues) {
//get the radius
var radius = calcPropRadius(circleValues[key]);
$("#" + key).attr({
cy: 59 - radius,
r: radius,
});
$("#" + key + "-text").text(
Math.round(circleValues[key] * 100) / 100 + " million"
);
}
}
Dont forget to add updateLegend(attribute) to the end of updatePropSymbols.
Activity 6 is all based on MegaCities.geojson, you don't need to use NYC Taxi Dataset or your own datasets.
If you have trouble with Chapter 5, you can download completed version of the folder from Chapter 05 Solutions and start Chapter 6 on the clean folder.
- Refactor any duplicate code in your script using either procedural or object-oriennted programming.
- Create a temporal or sequence legend on your leaflet map.
- Create an attribute legend for your Leaflet map based on the minimum, mean, and maximum attribute values across all attributes in your dataset. (You should add a updateLegend() function to update the attribute legend when click the buttons)
//PSEUDOCODE FOR UPDATE ATTRIBUTE LEGEND
Step 1. Update the content for legend (e.g. Population in XXXX)
Step 2. Update `<circle>` element for each of three attribute values: min, max, and mean
You may might need a separate function to get min, max, and mean from the dataset
of the current layer
The final result should look like this, and pay attention to the legend on the botton-right, which is updated for each temporal point.
- Zip a copy of your unit-2 folder and upload it to Canvas. The folder should include the
index.html,main.jsin thejsfolder,style.cssin thecssfolder, leaflet and jQuery library in yourlibfolder,MegaCities.geojsonin thedatafolder, and two images in theimgfolder. - Optional: Commit the changes to your unit-2 directory and sync with GitHub. Include "Activity 6" in the commit message (Summary).
This work is licensed under a Creative Commons Attribution 4.0 International License.
For more information, please contact Robert E. Roth ([email protected]).