At the moment, we show axis labels to one decimal place.
However, this breaks down when we have intervals such as 0.25.
We should therefore implement an algorithm for axis label rounding.
This is more complex than it seems due to floating point error, and not all decimal places being significant in conveying information.
Ideally, we could use the number of decimal places provided by the step, but this has flow on effects to font measuring.
At the moment, we show axis labels to one decimal place. However, this breaks down when we have intervals such as 0.25.
We should therefore implement an algorithm for axis label rounding.
This is more complex than it seems due to floating point error, and not all decimal places being significant in conveying information. Ideally, we could use the number of decimal places provided by the
step
, but this has flow on effects to font measuring.