Here is my code:
import matplotlib.pyplot as plt
plt.loglog(length,time,'--')
where length and time are lists.
How do I find the slope of this graph?
views:
233answers:
1
+6
A:
If you have matplotlib then I believe you must also have numpy installed. If that's true, then you could use numpy.polyfit to find the slope:
import matplotlib.pyplot as plt
import numpy as np
length=np.random.random(10)
length.sort()
time=np.random.random(10)
time.sort()
slope,intercept=np.polyfit(np.log(length),np.log(time),1)
print(slope)
plt.loglog(length,time,'--')
plt.show()
unutbu
2010-02-11 12:46:03
This would be true if the graph is a line. In case it is a curve and the slope changes at different points a diffrential is needed. Try the diff function.
whatnick
2010-02-11 13:21:43
Can you please provide some more details about the function.
Bruce
2010-02-12 07:05:02
@Peter: `polyfit` (in its simplest incarnation) takes 3 args: the `x`-data, `y`-data, and the degree of polynomial. Since you are looking for a linear fit, the 3rd arg is set to 1. `polyfit` then returns the coefficients of the best-fit polynomial, which in this case means the slope and y-intercept. HTH.
unutbu
2010-02-12 19:37:27