How to use SciPy to curve fit in Python || Python for Engineers

Curve fitting in Python is accomplished using Scipy.optimize.curve_fit. Curve_fit requires the user to define a function for the general form of the fit. If a linear fit is desired, define a linear equation in your function of choice. For quadratic fits, define a quadratic equation. The idea is that Python will optimize the iterative process of fitting the function's variables. The constants are then indexed using constants[0][i], where i is the ith value of the constant. The list constants[0] contains all the constants, so each constant must be indexed individually afterwards.
Packages required:
Matplotlib
Scipy
Numpy
GitHub source code: github.com/hendog993/PyNginee...
Linkedin: / henry-gilbert
Website: henrygilbert609.wixsite.com/h...

[@_Noolan_]

This is a fine video. Maybe not only one you'd like to watch for curve_fit but it shows you how to think of linear fit

[@hp1612]

Thank you so much! You don't know how much you had helped me there

[@laltugazi6372]

Thank you very much .a good understanding

[@mohamadachour4125]

you made me happy, Thank you

[@axelsenn920]

Thank you!

[@harshalbhoir8986]

too good bro

[@TheRavemaster28]

Hey, Very nice video and easy to understand. My Question: Is there a function that would give me the Mean Average Error of the function regarding to my Dataset?

[@An555gelo]

safed my life

[@benlaroche2777]

referring to: 4:00. What is at the other indices of "constants"? For example, what does "constants[1]" and "constants[2]" return?

[@rohitborooah4695]

Is there a method to fit my data ponts with the curve of best fit? That is I do not want to define my function, but let the method handle it? I just want the equation of the best fitting curve.

[@tonycardinal413]

Thank you so much for posting this. Ques: The curve_fit function returns the covariances between the parameters (coefficients for x and the intercept, etc) If there is a high covariance between the parameters what does this tell us? Is this bad? If so why? thanx!

[@arisoda]

6:10 so how can you get the formula of that trend-line? The a and b are obviously not the ones in ax+b, so how do I get this formula?

[@joebasswizard]

he took a slight shortcut here. Basically, you approximate the trend-line shape (here, logarithmic), plug in the appropriate formula, then let SciPy do the rest. The formula chosen is the best guess for this graph shape (a * log(x) + b), where (crudely) 'a' will affect the steepness of the curve and 'b' will affect the y axis offset. The a and b values are left entirely up to SciPy to calculate.

[@hanihazzouri275]

@@joebasswizard Hello from the future. The log function displayed in this example is relatively on the simple side, but some line formulas can get really complicated. Is there a step or code you can take to find that formula or do I need to call my math wiz friends every time?