I have been using d3 to create a multiline chart with focus and context brushing. Everything is going well except on the transition the dots at the data points with the tooltips are moving to a completely wrong position. I can't figure out what is causing this. Any help would be much appreciated. I attached the full code here and noted on the graph where I'm pretty sure the bug should be:
http://jsbin.com/osumaq/20/edit
When the button is clicked, a new json is passed to the graph to read.
The buggy block of code I think is this:
topicEnter.append("g").selectAll(".dot")
.data(function (d) { return d.values }).enter().append("circle").attr("clip-path", "url(#clip)")
.attr("stroke", function (d) {
return color(this.parentNode.__data__.name)
})
.attr("cx", function (d) {
return x(d.date);
})
.attr("cy", function (d) {
return y(d.probability);
})
.attr("r", 5)
.attr("fill", "white").attr("fill-opacity", .5)
.attr("stroke-width", 2).on("mouseover", function (d) {
div.transition().duration(100).style("opacity", .9);
div.html(this.parentNode.__data__.name + "<br/>" + d.probability).style("left", (d3.event.pageX) + "px").style("top", (d3.event.pageY - 28) + "px").attr('r', 8);
d3.select(this).attr('r', 8)
}).on("mouseout", function (d) {
div.transition().duration(100).style("opacity", 0)
d3.select(this).attr('r', 5);
});
Thank you very much.
What do you mean by tooltip ? Is it the window that appears when we hover on dots ? They seem fine. What I can see is that your dots are not moving while the lines are, and if I had to guess I would say your enter and update selections are mixed. If the dots are already on screen and you want to update their position (by calling your method update) you should have somthing along these lines :
// Bind your data
topicEnter.append("g").selectAll(".dot")
.data(function (d) { return d.values })
// Enter selection
topicEnter.enter().append("circle").attr("clip-path", "url(#clip)").attr("class", "dot");
// Update all the dots
topicEnter.attr("stroke", function (d) {
return color(this.parentNode.__data__.name)
})
.attr("cx", function (d) {
return x(d.date);
})
.attr("cy", function (d) {
return y(d.probability);
})
[...]
Related
guys, I am working with the most recent beeswarm visualization by Mike Bostock.
I wonder, in D3js v.4, how to change the color of one single specific point. In this case, I am talking about unemployment in more than 180 countries, and I would like to single out only "Brazil" with a different color.
How can I pick Brazil's dot and make it with a different color? As last resort, I suppose I could use SVG to single it out, but its to big a code, burden. I am fairly new to D3.
Here is the code: http://codepen.io/voltdatalab/pen/KzrNGo
var cell = g.append("g")
.attr("class", "cells")
.selectAll("g").data(d3.voronoi()
.extent([[-margin.left, -margin.top], [width + margin.right, height + margin.top]])
.x(function(d) { return d.x; })
.y(function(d) { return d.y; })
.polygons(data)).enter().append("g");
cell.append("circle")
.attr("r", 5)
.attr("cx", function(d) { return d.data.x; })
.attr("cy", function(d) { return d.data.y; })
;
Here is what I want to do:
I did this manipulating the SVG, not D3 code
You need to add .attr("fill",function(d) { if(d.data.id === "Djibouti")return "red"; }) to circle. Following is the change you need to make:
cell.append("circle")
.attr("r", 5)
.attr("cx", function(d) { return d.data.x; })
.attr("cy", function(d) { return d.data.y; })
.attr("fill",function(d) { if(d.data.id === "Djibouti")return "red"; })
;
Replace Djibouti with Brazil.
Here is a bin. Let me know if you want something like this.
Using D3 ver 3.5.5. I am using an example (https://gist.github.com/stepheneb/1182434) as a template: the example code to draw the data looks like this:
var circle = this.vis.select("svg").selectAll("circle")
.data(this.points, function(d) { return d; });
circle.enter().append("circle")
.attr("class", function(d) { return d === self.selected ? "selected" : null; })
.attr("cx", function(d) { return self.x(d.x); })
.attr("cy", function(d) { return self.y(d.y); })
.attr("r", 10.0)
.style("cursor", "ns-resize")
.on("mousedown.drag", self.datapoint_drag())
.on("touchstart.drag", self.datapoint_drag());
circle
.attr("class", function(d) { return d === self.selected ? "selected" : null; })
.attr("cx", function(d) {
return self.x(d.x); })
.attr("cy", function(d) { return self.y(d.y); });
circle.exit().remove();
I think of this as four sections: the first does selectAll("circles") and adds the data. The second tells where the data points are ("cx", "cy") and other attr(), and the third is a bit of mystery to me, because it appears to also set "cx" and "cy", but no other attributes. Finally, we do and exit().remove(), which the documentation says removes any data elements not associated with the data array. I dont see how this is happening in this example. When I set breakpoints into the code, both the "cx" steps get called for each data point in the this.points array.
In my code, I try to do the same steps:
hr_circles = self.graph_gps.svg.selectAll("hr_circles")
.data(self.graph_gps.datay1); // , function(d){return d;}
hr_circles.enter().append("circle")
.style("z-index", 3)
.attr("class", "y1")
.attr("r", 1)
.attr("cx", function (d, i) {
return xScale(d.time)
})
.attr("cy", function (d, i) {
return yScale(d.vy)
})
.on("mouseover",
function (d) {...displays a tooltip...})
.on("mouseout", function (d) {
});
hr_circles.attr("class", "y1")
.attr("cx", function (d, i) {
return xScale(d.time)
})
.attr("cy", function (d, i) {
return yScale(d.vy)
})
hr_circles.exit().remove();
When my graph initially displays, the data appear just fine, properly scaled, etc. When I try to re-scale by dragging on the x-axis (as in the example), the axis rescales itself just fine, and re-scaled data appears on the graph, but the original data is also still there (no longer scaled correctly), making a big mess! How do you erase or make the originally scaled data go away?
Tried to post images, but I guess my reputation is too low. Will send to anyone interested.
I have this visualization and I'm trying to add fisheye view to the chart. I have tried adding it with the following lines in the plotData function but it doesn't happen:
var fisheye = d3.fisheye.circular()
.radius(120);
svg.on("mousemove", function () {
fisheye.focus(d3.mouse(this));
circle.each(function (d) {
d.fisheye = fisheye(d);
});
});
Any ideas on how to solve this?
Thanks!
First things first, your d3.timer never stops running. This is driving my machine crazy (cpu 100%) and killing the performance of the fishey. I'm really not sure what you are doing there, so ignoring that for a moment.
Your fisheye needs a little massaging. First, it expects your data pixel's positions to be stored in d.x and d.y attributes. You can fudge this in with when drawing your circles:
circle
.attr("cx", function(d, i) { d.x = X(d[0]); return d.x; })
.attr("cy", function(d, i){ d.y = Y(d[1]); return d.y; });
Second, you are plotting your data in multiple steps, so you need to select all the circles for the fisheye. And third, you forgot the code that actually makes the points grow and shrink:
svg.on("mousemove", function () {
fisheye.focus(d3.mouse(this));
// select all the circles
d3.selectAll("circle.data").each(function(d) { d.fisheye = fisheye(d); })
// make them grow and shrink and dance
.attr("cx", function(d) { return d.fisheye.x; })
.attr("cy", function(d) { return d.fisheye.y; })
.attr("r", function(d) { return d.fisheye.z * 4.5; });
});
Updated example.
I would like to reproduce the process from D3 Sankey chart using circle node instead of rectangle node, however, I would like to select only certain nodes to change from rectangles to circles.
For example, in this jsfiddle used in the example, how would you only select Node 4 and Node 7 to be converted to a circle?
I updated your fiddle.
Basically you just need some way to select the nodes that you want to make different. I used unique classname so that you can style them with CSS as well. I didn't feel like writing the code to select just 4 and 7 (I'm lazy) so I just selected all of the even nodes instead.
// add in the nodes
var node = svg.append("g").selectAll(".node")
.data(graph.nodes)
.enter().append("g")
.attr("class", function (d, i) { return i % 2 ? "node rect" : "node circle";
})
Then you can use that to select the nodes and add circles instead of rectangles.
svg.selectAll(".node.circle").append("circle")
.attr("r", sankey.nodeWidth() / 2)
.attr("cy", function(d) { return d.dy/2; })
.attr("cx", sankey.nodeWidth() / 2)
.style("fill", function (d) {
There is also another similar approach, illustrated in the following jsfiddle.
I started from this fiddle (from another SO question that you merntioned)), where all nodes had already been converted to circles:
Then I modified existing and added some new code that involves filtering during creation of circles:
// add the circles for "node4" and "node7"
node
.filter(function(d){ return (d.name == "node4") || (d.name == "node7"); })
.append("circle")
.attr("cx", sankey.nodeWidth()/2)
.attr("cy", function (d) {
return d.dy/2;
})
.attr("r", function (d) {
return Math.sqrt(d.dy);
})
.style("fill", function (d) {
return d.color = color(d.name.replace(/ .*/, ""));
})
.style("fill-opacity", ".9")
.style("shape-rendering", "crispEdges")
.style("stroke", function (d) {
return d3.rgb(d.color).darker(2);
})
.append("title")
.text(function (d) {
return d.name + "\n" + format(d.value);
});
// add the rectangles for the rest of the nodes
node
.filter(function(d){ return !((d.name == "node4") || (d.name == "node7")); })
.append("rect")
.attr("y", function (d) {
return d.dy/2 - Math.sqrt(d.dy)/2;
})
.attr("height", function (d) {
return Math.sqrt(d.dy);
})
.attr("width", sankey.nodeWidth())
.style("fill", function (d) {
return d.color = color(d.name.replace(/ .*/, ""));
})
.style("fill-opacity", ".9")
.style("shape-rendering", "crispEdges")
.style("stroke", function (d) {
return d3.rgb(d.color).darker(2);
})
.append("title")
.text(function (d) {
return d.name + "\n" + format(d.value);
});
Similar code had to be modified for accurate positioning text beside rectangles.
I believe the final result looks natural, even though it lost some of the qualities of the original Sankey (like wider connections).
I've got a sortable heat map that I've created in D3 shown here: http://bl.ocks.org/umcrcooke/5703304
When I click on the year (column) the initial sort/transition works well, but subsequent clicks resorts, but without the transition. I'm having difficulty troubleshooting it. The code for the transition listed below:
I've set it up such that when the column text is clicked the update function executes:
.on("click", function(d,i) { return d3.transition().each(update(d));});
And the relevant pieces of the update function are:
function update(year) {
grid.selectAll('rect')
.transition()
.duration(2500)
.attr("y", function(d) { return (sortOrder[year].indexOf(d.Country))*cell.height; })
grid.selectAll(".cell_label")
.transition()
.duration(2500)
.attr("y", function(d) { return (sortOrder[year].indexOf(d.Country))*cell.height + (cell.height-cell.border)/2; })
d3.selectAll(".row_label")
.sort(function(a, b) {
return d3.ascending(+a[year], +b[year]);
})
.transition()
.duration(2500)
.attr("y", function(d, i) { return (i*cell.height) + (cell.height-cell.border)/2; });
}
I'm not sure what you're trying to do with d3.transition().each() in the handler, but you don't need it. Changing to:
.on("click", function(d,i) { update(d); });
fixes the problem. See fiddle: http://jsfiddle.net/nrabinowitz/Lk5Pw/