If for example
L := [3, 4, 6, 7, 2, 3, 5, 4, 6, 8, 20, 4, 5, 12, 0, 5, 5, 5, 3]:
b := ListTools:-Enumerate(L):
g := plot(CurveFitting:-Spline(b, v), v = 1 .. 19):
gp:=pointplot(b):
display(g,gp)
It is a nice curve but between points at x=12 and x=13 there exists a large dip, and there are a few other smaller ones. But of course Spline uses a piecwise polynomial so I understand the overshoots but can we not fit an even better curved line? So that point 12 is a point of concavity? If we can how do we do it?
