- When to use a line chart:
To track rises and falls over time
To reveal patterns in the data: trends, fluctuations, cycles, rates of change
When the labels on the x-axis have a natural ordering (2005, 2006, 2007...)
- When NOT to use a line chart:
If your data is categorized without a natural ordering, for example by country, department, or activity, a
chart works better
It can often be a challenge, however, to display a lot of trends on one line chart clearly and succinctly.
- With Bime there are a few easy ways to display a vast amount of data on a line chart, in a simple and concise way.
Filter your results
To quote one of my favourite acronyms, KISS - Keep it simple, stupid! Only show the metrics which you need to display and are relevant to your analysis. In Bime there are several easy ways to do this. Firstly, by clicking the drop down arrow next to your attribute on the frame and selecting or deselecting the elements you wish to display. After executing the calculation you can again select which elements you wish to show by clicking them in the row selector on the left of the chart (hold 'ctrl' or 'cmd' to select multiple elements and 'shift' to select a list of elements).
Alternatively, you can filter your measures by only showing the top or bottom x number of results using the 'Top' tab of the post-processing menu.
Highlight a metric
Another quick technique is to simply highlight a measure using the mouse. Just hover the cursor over the name of an element in the legend on the right and that element will be shown while the others are faded out. For example in the image below we are able to just show the values for 'Internet Explorer'.
Cycle through your results
In Bime you can switch through your selected row members automatically, simply click the 'play' button above the list of row members or manually click through.
Group your attribute values
To make a chart less cluttered, group similar values. For example in the initial image in this post we had a visualization looking at the web browsers used to access a website, in which we had several different Blackberry browsers which can obviously be grouped together. To group attribute elements; click the arrow next to the attribute you wish to make a group from, select 'create a group' to bring up the group selection box and then follow the instructions in the images below.
Show several measures simultaneously
In Bime it is possible to display several measures on a line chart, in a succinct way, to maintain the simplicity of the visualization, even when the values are far apart numerically. The three examples below show both visitors per browser and visit bounce rate. Bime offers 3 options:
[one_third]1. Dual Axis - Display two different measures with two quantitative scales on one chart, regardless of the units.[/one_third]
[one_third]2. Size - Display your first measure on the axis and your second measure using the size (thickness) of the line. The visualization again shows visitors per browser and visit bounce rate.[/one_third]
[one_third]3. Color - Display your first measure on the axis and your second measure using color. In this visualization we can see that the spike in Chrome, a large number of visitors that month, also corresponded with a high bounce rate.
TIP: In this last visualization the 'curve' option has been selected.[/one_third]