First and foremost, before you get started, you’ll need to
identify what data you’d like to represent in your chart. You may need to whittle down your data so that it
isn’t overwhelming but still fairly represents your population. For this
purpose, you might consider using a Filter node to filter down to just the
entries you’re truly interested in, or using a Cleanse node to create an
“other” category to concatenate some of the smaller categories.
Once you have the dataset you’d like to use and have opened
the chart node, your first step will be to select the chart type (pie chart,
bar chart, etc.). Once you’ve done that, you’ll want to fill in all necessary
fields on the right side of the Chart Data window, such as deciding which
variables should be your x and y axes. Once everything has been labeled, you
can edit the look of your chart by clicking on the colored icon in the top
right of the window.
As you begin to navigate through the Chart Style Editor
window, here are some things to keep in mind:
1. The chart type does not automatically identify the chart
type that was selected in the previous window – you’ll want to update this and
select the appropriate chart type.
2. The sample chart you see on the right will update to reflect
any changes you make in the chart editor. Use it to evaluate your aesthetic
decisions as you go.
3. Below the sample chart you can choose whether or not to
include data labels and adjust the font, size, style, color, and chart
background.
4. When charting binary or categorical variables, you may not
see a need to include a legend. If you select the ‘Legend’ tab and uncheck the
‘Display Legend’ box, you can remove the legend. If you do remove the legend,
be sure the chart is labeled accurately in the first window.
5. If you’d like to make any of your charts 3D, you can do so
by going to the ‘3D’ tab and checking the ‘Display in 3D’ box. There you can
also change the inclination, depth, rotation, etc.
6. Once you’ve edited the background color and text color of
your chart in (below the sample chart), you can choose a color palette for the
rest of the chart by going to the ‘Color Palette’ tab. There you can select a
pre-programmed palette or create your own custom palette.
7. One of the more subtle functions in the chart node is the
ability to shade ranges within a chart – found in the ‘Axis’ tab. Here you can
choose to assign colors to ranges within a chart to better visualize certain
areas of interest in your data. This is especially nice when looking at
retention rates or variables represented as percentages because the ranges are
easy to define & interpret.
8. Another use of the ‘Axis’ tab is to help label your data.
Too many times I’ve gone to chart my data but found that half my data points
are unlabeled in my graph. Here, choose the axis that should contain those
missing labels and unselect the ‘Auto Label’ box. Now under the ‘Major Tick
Marks’ heading, unselect the ‘Auto Interval’ box and set the interval to be 1
to insert all missing data point labels. It should look something like this (note
that I’ve chosen the y-axis):
9. Now that you know how to edit each chart individually, you
might find that you like a particular style and want to save it to use later.
You can do this within the Chart Node by clicking on the ‘…’ button next to the
‘Chart Style’ drop-down. Here, click the green plus button to add a new chart
style – give it a name and you’ll find yourself in the ‘Chart Style Editor’
window. Once you’ve finished editing, you’ll be able to access your saved chart
style within the Chart Node from the Chart Style drop-down menu.
10. Once you’ve created a chart type that you love, you can also
export it to the Collaborative Cloud so that others can use it – if you have a
minute, check out the cloud and search for chart styles for the opportunity to
download styles that other users have created.
Do you have any tips for charting, or questions about how to
use the charting node? Leave them in the comments below :)
Happy charting!
-by Jon MacMillan, Data Analyst at Rapid Insight
No comments:
Post a Comment