The same fit as the logarithmic example, on an axis where the curve is visible
The fit is y = a × e^(b×x). On a logarithmic y axis an
exponential is a straight line, so a trendline drawn as one segment between the
first and last point is indistinguishable from the curve. On a linear
axis it is not: 2^x over x = 0–10 is 1024 at the end, but only 32 at
x = 5, where a straight chord would pass at 512.
The plugin samples the fitted curve, so the line below follows the data on either axis. Compare it with the logarithmic version, where the same fit appears as a straight line.
y = a × e^(b×x) instead of a straight regressiontrue — the label shows the fitted a and b"solid"