Create a gist now

Instantly share code, notes, and snippets.

Topic Clouds using D3 Word Cloud Layout
// Word cloud layout by Jason Davies, http://www.jasondavies.com/word-cloud/
// Algorithm due to Jonathan Feinberg, http://static.mrfeinberg.com/bv_ch03.pdf
(function(exports) {
function cloud() {
var size = [256, 256],
text = cloudText,
font = cloudFont,
fontSize = cloudFontSize,
rotate = cloudRotate,
padding = cloudPadding,
spiral = archimedeanSpiral,
words = [],
timeInterval = Infinity,
event = d3.dispatch("word", "end"),
timer = null,
cloud = {};
cloud.start = function() {
var board = zeroArray((size[0] >> 5) * size[1]),
bounds = null,
n = words.length,
i = -1,
tags = [],
data = words.map(function(d, i) {
return {
text: text.call(this, d, i),
font: font.call(this, d, i),
rotate: rotate.call(this, d, i),
size: ~~fontSize.call(this, d, i),
padding: cloudPadding.call(this, d, i)
};
}).sort(function(a, b) { return b.size - a.size; });
if (timer) clearInterval(timer);
timer = setInterval(step, 0);
step();
return cloud;
function step() {
var start = +new Date,
d;
while (+new Date - start < timeInterval && ++i < n && timer) {
d = data[i];
d.x = (size[0] * (Math.random() + .5)) >> 1;
d.y = (size[1] * (Math.random() + .5)) >> 1;
cloudSprite(d, data, i);
if (place(board, d, bounds)) {
tags.push(d);
event.word(d);
if (bounds) cloudBounds(bounds, d);
else bounds = [{x: d.x + d.x0, y: d.y + d.y0}, {x: d.x + d.x1, y: d.y + d.y1}];
// Temporary hack
d.x -= size[0] >> 1;
d.y -= size[1] >> 1;
}
}
if (i >= n) {
cloud.stop();
event.end(tags, bounds);
}
}
}
cloud.stop = function() {
if (timer) {
clearInterval(timer);
timer = null;
}
return cloud;
};
cloud.timeInterval = function(x) {
if (!arguments.length) return timeInterval;
timeInterval = x == null ? Infinity : x;
return cloud;
};
function place(board, tag, bounds) {
var perimeter = [{x: 0, y: 0}, {x: size[0], y: size[1]}],
startX = tag.x,
startY = tag.y,
maxDelta = Math.sqrt(size[0] * size[0] + size[1] * size[1]),
s = spiral(size),
dt = Math.random() < .5 ? 1 : -1,
t = -dt,
dxdy,
dx,
dy;
while (dxdy = s(t += dt)) {
dx = ~~dxdy[0];
dy = ~~dxdy[1];
if (Math.min(dx, dy) > maxDelta) break;
tag.x = startX + dx;
tag.y = startY + dy;
if (tag.x + tag.x0 < 0 || tag.y + tag.y0 < 0 ||
tag.x + tag.x1 > size[0] || tag.y + tag.y1 > size[1]) continue;
// TODO only check for collisions within current bounds.
if (!bounds || !cloudCollide(tag, board, size[0])) {
if (!bounds || collideRects(tag, bounds)) {
var sprite = tag.sprite,
w = tag.width >> 5,
sw = size[0] >> 5,
lx = tag.x - (w << 4),
sx = lx & 0x7f,
msx = 32 - sx,
h = tag.y1 - tag.y0,
x = (tag.y + tag.y0) * sw + (lx >> 5),
last;
for (var j = 0; j < h; j++) {
last = 0;
for (var i = 0; i <= w; i++) {
board[x + i] |= (last << msx) | (i < w ? (last = sprite[j * w + i]) >>> sx : 0);
}
x += sw;
}
delete tag.sprite;
return true;
}
}
}
return false;
}
cloud.words = function(x) {
if (!arguments.length) return words;
words = x;
return cloud;
};
cloud.size = function(x) {
if (!arguments.length) return size;
size = [+x[0], +x[1]];
return cloud;
};
cloud.font = function(x) {
if (!arguments.length) return font;
font = d3.functor(x);
return cloud;
};
cloud.rotate = function(x) {
if (!arguments.length) return rotate;
rotate = d3.functor(x);
return cloud;
};
cloud.text = function(x) {
if (!arguments.length) return text;
text = d3.functor(x);
return cloud;
};
cloud.spiral = function(x) {
if (!arguments.length) return spiral;
spiral = spirals[x + ""] || x;
return cloud;
};
cloud.fontSize = function(x) {
if (!arguments.length) return fontSize;
fontSize = d3.functor(x);
return cloud;
};
cloud.padding = function(x) {
if (!arguments.length) return padding;
padding = d3.functor(x);
return cloud;
};
return d3.rebind(cloud, event, "on");
}
function cloudText(d) {
return d.text;
}
function cloudFont() {
return "serif";
}
function cloudFontSize(d) {
return Math.sqrt(d.value);
}
function cloudRotate() {
return (~~(Math.random() * 6) - 3) * 30;
}
function cloudPadding() {
return 1;
}
// Fetches a monochrome sprite bitmap for the specified text.
// Load in batches for speed.
function cloudSprite(d, data, di) {
if (d.sprite) return;
c.clearRect(0, 0, (cw << 5) / ratio, ch / ratio);
var x = 0,
y = 0,
maxh = 0,
n = data.length;
di--;
while (++di < n) {
d = data[di];
c.save();
c.font = ~~((d.size + 1) / ratio) + "px " + d.font;
var w = c.measureText(d.text + "m").width * ratio,
h = d.size << 1;
if (d.rotate) {
var sr = Math.sin(d.rotate * cloudRadians),
cr = Math.cos(d.rotate * cloudRadians),
wcr = w * cr,
wsr = w * sr,
hcr = h * cr,
hsr = h * sr;
w = (Math.max(Math.abs(wcr + hsr), Math.abs(wcr - hsr)) + 0x1f) >> 5 << 5;
h = ~~Math.max(Math.abs(wsr + hcr), Math.abs(wsr - hcr));
} else {
w = (w + 0x1f) >> 5 << 5;
}
if (h > maxh) maxh = h;
if (x + w >= (cw << 5)) {
x = 0;
y += maxh;
maxh = 0;
}
if (y + h >= ch) break;
c.translate((x + (w >> 1)) / ratio, (y + (h >> 1)) / ratio);
if (d.rotate) c.rotate(d.rotate * cloudRadians);
c.fillText(d.text, 0, 0);
c.restore();
d.width = w;
d.height = h;
d.xoff = x;
d.yoff = y;
d.x1 = w >> 1;
d.y1 = h >> 1;
d.x0 = -d.x1;
d.y0 = -d.y1;
x += w;
}
var pixels = c.getImageData(0, 0, (cw << 5) / ratio, ch / ratio).data,
sprite = [];
while (--di >= 0) {
d = data[di];
var w = d.width,
w32 = w >> 5,
h = d.y1 - d.y0,
p = d.padding;
// Zero the buffer
for (var i = 0; i < h * w32; i++) sprite[i] = 0;
x = d.xoff;
if (x == null) return;
y = d.yoff;
var seen = 0,
seenRow = -1;
for (var j = 0; j < h; j++) {
for (var i = 0; i < w; i++) {
var k = w32 * j + (i >> 5),
m = pixels[((y + j) * (cw << 5) + (x + i)) << 2] ? 1 << (31 - (i % 32)) : 0;
if (p) {
if (j) sprite[k - w32] |= m;
if (j < w - 1) sprite[k + w32] |= m;
m |= (m << 1) | (m >> 1);
}
sprite[k] |= m;
seen |= m;
}
if (seen) seenRow = j;
else {
d.y0++;
h--;
j--;
y++;
}
}
d.y1 = d.y0 + seenRow;
d.sprite = sprite.slice(0, (d.y1 - d.y0) * w32);
}
}
// Use mask-based collision detection.
function cloudCollide(tag, board, sw) {
sw >>= 5;
var sprite = tag.sprite,
w = tag.width >> 5,
lx = tag.x - (w << 4),
sx = lx & 0x7f,
msx = 32 - sx,
h = tag.y1 - tag.y0,
x = (tag.y + tag.y0) * sw + (lx >> 5),
last;
for (var j = 0; j < h; j++) {
last = 0;
for (var i = 0; i <= w; i++) {
if (((last << msx) | (i < w ? (last = sprite[j * w + i]) >>> sx : 0))
& board[x + i]) return true;
}
x += sw;
}
return false;
}
function cloudBounds(bounds, d) {
var b0 = bounds[0],
b1 = bounds[1];
if (d.x + d.x0 < b0.x) b0.x = d.x + d.x0;
if (d.y + d.y0 < b0.y) b0.y = d.y + d.y0;
if (d.x + d.x1 > b1.x) b1.x = d.x + d.x1;
if (d.y + d.y1 > b1.y) b1.y = d.y + d.y1;
}
function collideRects(a, b) {
return a.x + a.x1 > b[0].x && a.x + a.x0 < b[1].x && a.y + a.y1 > b[0].y && a.y + a.y0 < b[1].y;
}
function archimedeanSpiral(size) {
var e = size[0] / size[1];
return function(t) {
return [e * (t *= .1) * Math.cos(t), t * Math.sin(t)];
};
}
function rectangularSpiral(size) {
var dy = 4,
dx = dy * size[0] / size[1],
x = 0,
y = 0;
return function(t) {
var sign = t < 0 ? -1 : 1;
// See triangular numbers: T_n = n * (n + 1) / 2.
switch ((Math.sqrt(1 + 4 * sign * t) - sign) & 3) {
case 0: x += dx; break;
case 1: y += dy; break;
case 2: x -= dx; break;
default: y -= dy; break;
}
return [x, y];
};
}
// TODO reuse arrays?
function zeroArray(n) {
var a = [],
i = -1;
while (++i < n) a[i] = 0;
return a;
}
var cloudRadians = Math.PI / 180,
cw = 1 << 11 >> 5,
ch = 1 << 11,
canvas,
ratio = 1;
if (typeof document !== "undefined") {
canvas = document.createElement("canvas");
canvas.width = 1;
canvas.height = 1;
ratio = Math.sqrt(canvas.getContext("2d").getImageData(0, 0, 1, 1).data.length >> 2);
canvas.width = (cw << 5) / ratio;
canvas.height = ch / ratio;
} else {
// node-canvas support
var Canvas = require("canvas");
canvas = new Canvas(cw << 5, ch);
}
var c = canvas.getContext("2d"),
spirals = {
archimedean: archimedeanSpiral,
rectangular: rectangularSpiral
};
c.fillStyle = "red";
c.textAlign = "center";
exports.cloud = cloud;
})(typeof exports === "undefined" ? d3.layout || (d3.layout = {}) : exports);
<html>
<head>
<title>Topic Clouds</title>
<meta http-equiv="Content-Type" content="text/html;charset=utf-8"/>
<script src="http://d3js.org/d3.v2.min.js?2.10.0"></script>
<script type="text/javascript" src="d3.layout.cloud.js"></script>
</head>
<body>
<script>
wordScale=d3.scale.linear().domain([1,100,1000,10000]).range([10,20,40,80]).clamp(true);
wordColor=d3.scale.linear().domain([10,20,40,80]).range(["blue","green","orange","red"]);
for (x = 0; x < 3; x++) {
viz = d3.select("body").append("svg")
.attr("width", 400)
.attr("height", 440)
.attr("id", "svg" + x);
}
for (x = 0; x < 3; x++) {
//d3.json("topic.json", function(topic) {
d3.csv("topic_words"+x+".csv", function(topic) {
d3.layout.cloud().size([400, 400])
// .words([{"text":"test","size":wordScale(.01)},{"text":"bad","size":wordScale(1)}])
.words(topic)
.rotate(function() { return ~~(Math.random() * 2) * 5; })
.fontSize(function(d) { return wordScale(d.size); })
.on("end", draw)
.start();
function draw(words) {
viz = d3.select("#svg" + topic[0].topic);
viz.append("g")
.attr("transform", "translate(200,220)")
.selectAll("text")
.data(words)
.enter().append("text")
.style("font-size", function(d) { return d.size + "px"; })
.style("fill", function(d) { return wordColor(d.size); })
.style("opacity", .75)
.attr("text-anchor", "middle")
.attr("transform", function(d) {
return "translate(" + [d.x, d.y] + ")rotate(" + d.rotate + ")";
})
.text(function(d) { return d.text; });
viz
.append("text")
.data([topic[0]])
.style("font-size", 20)
.style("font-weight", 900)
.attr("x", 100)
.attr("y", 20)
.text(function(d) { return "TOPIC " + d.topic; })
// d3.select("#svg"+x).append("svg:text").text("Topic " + x);
// viz.enter().append("svg:text").text("Topic " + x);
}
})
}
</script>
text size topic access 1238 0 streets 1020 0 transportation 982 0 system 824 0 pedestrian 767 0 provide 763 0 bicycle 719 0 major 696 0 arterial 596 0 local 452 0 centers 415 0 service 388 0 connections 373 0 adjacent 337 0 arterials 335 0 facilities 317 0 modes 312 0 design 310 0 high 305 0 travel 284 0 areas 282 0 regional 275 0 serve 269 0 circulation 257 0 safe 241 0 vehicle 233 0 demand 228 0 roadway 223 0 improvements 218 0 corridors 211 0 capacity 208 0 activity 208 0 roadways 207 0 collector 202 0 modal 198 0 automobile 185 0 oriented 183 0 function 182 0 locations 182 0 light 174 0 convenient 172 0 trails 169 0 employment 156 0 classification 151 0 paths 151 0 designed 148 0 accommodate 136 0 primary 135 0 direct 135 0 type 131 0 wide 130 0 bicyclists 129 0 map 129 0 destinations 129 0 intersections 129 0 users 128 0 bikeways 126 0 complete 126 0 trail 120 0 bus 117 0 long 114 0 effective 113 0 providing 111 0 volumes 111 0 includes 110 0 standards 99 0 points 99 0 ride 99 0 truck 97 0 developments 97 0 emergency 94 0 bicycling 92 0 collectors 90 0 develop 86 0 motor 85 0 hours 80 0 additional 80 0 boulevard 79 0 balance 78 0 ridership 74 0 address 74 0 coordinate 72 0 serving 72 0 significant 72 0 stops 71 0 citywide 70 0 recreational 70 0 multiple 70 0 freeway 68 0 separated 68 0 minor 68 0 principal 68 0 highways 68 0 connect 66 0 amenities 66 0 landscaping 65 0 low 65 0 crossings 64 0 stop 64 0 sidewalks 63 0
text size topic buildings 460 1 plan 451 1 policy 442 1 neighborhoods 327 1 civic 301 1 community 249 1 strategies 235 1 existing 222 1 lots 221 1 walkable 217 1 large 207 1 arroyos 203 1 concerns 196 1 map 183 1 goal 182 1 spaces 182 1 neighborhood 182 1 streets 180 1 network 137 1 regional 136 1 addressing 131 1 compact 123 1 historic 110 1 element 105 1 valley 105 1 center 103 1 central 102 1 infill 101 1 small 95 1 space 92 1 greens 91 1 step 91 1 principles 91 1 design 89 1 blocks 88 1 drainage 87 1 drive 86 1 ave 82 1 smartcode 81 1 complete 81 1 impact 80 1 goals 80 1 legend 80 1 critical 79 1 plans 79 1 vacant 75 1 cities 74 1 current 74 1 green 72 1 sector 69 1 provided 68 1 arroyo 67 1 create 67 1 walk 67 1 proposed 65 1 fees 64 1 housing 64 1 preferred 64 1 northeast 64 1 war 63 1 eastside 63 1 officials 63 1 alleys 63 1 lined 62 1 world 61 1 mixed 59 1 created 59 1 smart 58 1 time 57 1 site 57 1 important 57 1 westside 57 1 surrounding 56 1 shed 56 1 lands 55 1 auto 54 1 homes 54 1 farmland 51 1 schools 51 1 subdivision 50 1 multiway 50 1 valuable 49 1 face 49 1 mission 49 1 owned 48 1 add 47 1 urbanism 47 1 shade 47 1 form 46 1 fabric 46 1 upper 46 1 added 46 1 sense 46 1 irrigated 45 1 title 45 1 regulations 45 1 edges 44 1 connected 43 1 fields 43 1 long 43 1
text size topic development 818 2 transit 746 2 centers 647 2 mixed 640 2 corridors 520 2 urban 443 2 city 414 2 oriented 388 2 building 359 2 areas 314 2 zoning 299 2 redevelopment 209 2 activity 198 2 blocks 195 2 districts 194 2 existing 190 2 vision 187 2 block 178 2 compact 166 2 investment 165 2 friendly 161 2 initiative 160 2 housing 158 2 standards 158 2 walking 154 2 develop 152 2 places 129 2 policies 127 2 people 125 2 stations 122 2 infill 117 2 options 114 2 sensitive 110 2 modal 108 2 walkable 106 2 design 105 2 quality 103 2 growth 103 2 creating 102 2 complete 101 2 key 99 2 private 91 2 homes 91 2 create 90 2 tod 88 2 encouraging 86 2 promotes 86 2 identify 83 2 concept 83 2 developments 83 2 mix 82 2 hubs 82 2 connected 81 2 direct 77 2 higher 77 2 vibrant 76 2 illustration 76 2 choices 75 2 built 74 2 includes 74 2 initiatives 73 2 core 68 2 corridor 67 2 scenario 65 2 jobs 64 2 shop 64 2 successful 64 2 type 62 2 principles 61 2 distance 59 2 practices 56 2 developer 53 2 destinations 53 2 exist 52 2 accommodate 51 2 sustainable 51 2 environment 51 2 central 50 2 patterns 49 2 developers 48 2 multifamily 48 2 future 47 2 pedestrian 47 2 small 46 2 investments 46 2 served 44 2 station 44 2 dense 42 2 trails 40 2 definition 40 2 similar 39 2 targeted 39 2 river 38 2 environments 38 2 natural 38 2 concentrations 36 2 walk 35 2 easy 35 2 university 35 2 boulevards 33 2
@codersofthedark
  1. It should be topic_words2.csv and not topic_words3.csv
  2. I am getting following error on running the code:
    Uncaught TypeError: Cannot read property 'topic' of undefined cloud.html:36
    draw cloud.html:36
    t d3.v2.min.js:1
    step d3.layout.cloud.js:60
    cloud.start d3.layout.cloud.js:36
    (anonymous function) cloud.html:32
    (anonymous function) d3.v2.min.js:1
    r d3.v2.min.js:2
    r.onreadystatechange d3.v2.min.js:2
@adhoch
adhoch commented Apr 9, 2014

The problem is that the csv's are all on one line. If you make each row a new line it will work fine.

@g1eb
g1eb commented Nov 22, 2016

That link to Jason Davies paper on word cloud layout has moved to https://www.jasondavies.com/wordcloud/

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment