Learn to visualize your data like an expert with these practical how-tos for presentation, analysis, and understanding.
Too many points to plot often means obscured patterns in the clutter. Density maps offer a smooth alternative.
The chart type often goes overlooked because people don't understand them. Maybe this will help.
As people and things move through a place, it can be useful to see their connected paths instead of just individual points.
A big part of statistics is comparisons, and perhaps more importantly, to figure out what to compare things to. Perspective changes with the baseline.
Text can provide much needed context to traditional visual cues and can be used as a visual cue itself in some cases.
A frequent challenge of visualization is behind the scenes, to get the data and to mold it into the format you need. Do that. Then map.
The combination of a time series chart and a scatter plot lets you compare two variables along with temporal changes.
Maybe you want to make spatial comparisons over time or across categories. Organized small maps might do the trick.
Also known as specialized or custom line charts. Figure out how to draw lines with the right spacing and pointed in the right direction, and you've got your slopegraphs.
Make a lot of charts at once, line them up in a grid, and you can make quick comparisons across several categories.
Although time series plots and small multiples can go a long way, animation can make your data feel more real and relatable. Here is how to do it in R via the animated GIF route.
When you plot a lot of data at once, points and lines can obscure others and hide patterns. Transparency can help reveal what is really there.
You get a lot of bang for the buck with R, charting-wise, but it can be confusing at first, especially if you've never written code. Here are some examples to get started.
When presented with a static graphic, it can be useful to see specific values after you see overall patterns. This tutorial shows you how to add simple interactions to a choropleth map so you can get specifics for regions.
You saw how to make basic heat maps a while back, but you might want more flexibility for a specific data set. Once you understand the components of a heat map, the rest is straightforward.
Sometimes these cartograms can distort areas beyond recognition, but they can also provide a better visual representation for a region with a wide range of subregions. At the least, they're fun to look at.
From the basic area chart, to the stacked version, to the streamgraph, the geometry is similar. Once you know how to do one, you can do them all.
When base graphics and existing packages don't do it for you, turn to low-level graphics functions to make what you want.
Single data points from a large dataset can make it more relatable, but those individual numbers don't mean much without something to compare to. That's where distributions come in.
These tend to be made ad hoc and are usually pieced together manually, which takes a lot of time. Here's a way to lay the framework in R, so you don't have to do all the work yourself.
Time series charts can easily turn to spaghetti when you have multiple categories. By highlighting the ones of interest, you can direct focus and allow comparisons.
The familiar but underused layout is a good way to look at patterns over time.