PyXPlot

# Functional forms to be fitted -- parabolas
f(x) = a * x**2 + b * x + c
g(x) = d * x**2 + e * x + f

# First of all, fit data neglecting errorbars
fit f() 'example8.dat' via a,b,c
# Now fit data taking errorbars into account
fit g() 'example8.dat' using 1:2:3 via d,e,f
# Now fit a spline through the data
spline h() 'example8.dat'

# Plot the resulting functions
set width 12
set key top xcentre
set xlabel 'x'
set ylabel 'y'
plot [0:8][0:5] \
     'example8.dat' with yerrorbars, f(x), g(x), h(x)

Notes

The fit command works in PyXPlot in essentially the same way as in gnuplot. In this example, we take a series of data points, and first fit parabolas through them. For the first fit, f(x), we do not take the errorbars into account; in the second, g(x), we do. Then, we use the spline command to fit a spline, h(x), through the same data. Strong oscillation is seen in this example because of the angular nature of the data; it is not well-fit by a spline.