Putting two graphs together

We often understand data more clearly as a picture. That is called data visualisation, and this section goes further with it.

Take the monthly onion prices in two towns. Yahapur runs 25, 24, 26, 28, 30, 35, 39, 43, 49, 56, 59, 44 through the year; Wahapur runs 19, 17, 23, 30, 38, 35, 42, 39, 53, 60, 52, 42.

You could draw two separate column graphs, one per town. But comparing them means looking back and forth between two pictures, and your eye cannot hold both at once.

So combine them: draw the bars side by side, one pair for each month. This is called a clustered column graph, and since there are two columns per cluster, a double column graph.

Two things make it readable. First, different colours separate the two towns. Second — and this is easy to overlook — the bars also carry dots and slanted lines. Those help people who find it difficult to distinguish colours, and they keep the graph readable when it is printed in greyscale.

Always identify the SCALE before reading any graph: how much does one unit of height stand for? Here the vertical axis is marked in tens of rupees.

Now the graph does the work. The relative heights of the two bars in each cluster tell you at a glance where onions are costlier that month, and you can estimate the DIFFERENCE by reading against the markings. Comparing month by month is far easier than it was with two separate graphs, and patterns over the year — both towns rising towards October, then falling — become visible at once.