Linear interpolation using scipy's interp1d

I am trying to do a linear interpolation

Code:

from scipy.interpolate import interp1d

x_ = [2.,   2.8, 16.7]
y_ = [0.87678869, 0.49044994, 1.4322675]
y_std = []
plt.plot(x_, y_, marker='o', color='k')
f = interp1d(x_, y_, kind='linear', bounds_error=False)

plt.plot(x_, f(x_), marker='o', color='tab:green')
plt.show()

I expected a linear line, but the result appears to be a piecewise fit. Could someone please suggest what's going wrong?

Interpolate just fills in the content between the points, and therefore it must hit every single point in your original dataset.

What you're looking for is to run a linear regression:

https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.linregress.html