missing data

In the hsteps style, empty lines, NaN values, and missing data have distinct meanings. If an empty line is present in the data, the data series is reset at that point. This is analogous to a blank line causing a break to start a new curve in the case of with lines. If an x-value contains NaN, it is processed in the same manner as an empty line. If the x-value is valid but the y-value contains NaN, no horizontal line is drawn for that particular data point but the x-value is still used if needed to estimate the step width.