SciPy curve_fit: What is "pcov"?
Vložit
- čas přidán 13. 07. 2024
- Check out my course on UDEMY: learn the skills you need for coding in STEM:
www.udemy.com/course/python-s...
In this video I show what errors on optimal parameters obtained using a curve fit actually mean. My video on curve fitting can be found here:
• Curve Fitting in Pytho...
Tutorial Playlist:
• The Full Python Tutorial
Code:
github.com/lukepolson/youtube...
Discord:
/ discord - Věda a technologie
wait..... you mean you're not a bird?
Lmfao
Nah, Billy just coded up a next level ML model that creates the human avatar you see
I'm learning about probability density functions (pdf) and cumulative distribution functions (cdf) this week in class. I didn't realize how applicable statistics is to programming.
It's really essential in for example state-of-the-art machine learning. For example, the commonly used SoftMax function is used to turn an arbitrary vector into a pd.
Great explanation. Also love the basic ones such as the class video from zero. Please do more of those!
Without any doubt, the best videos on scientific applications of Python. Thank you very much.
Great explanation! Just in time before a lab session where I’ll have to do nonlinear fitting
Great content! Hope that you'll continue making more of these videos :)
You should do a tutorial on Lmfit. It's really advanced and built on top curve fit.
Awesome, simply a bull's eye explanation
Can you please do a video on linear programming in python with PULP library !!!
It raised an error even for error bar graph saying: ValueError: 'yerr' must not contain negative values
Nice
1. what if using the KDE instead the χ^2 to find the pdf of a set of data, sometimes distributed differently and with some "outliers"?
2. Why not to use the sklearn to fit quite automatically (without specifying the curve fit equation ) a set of data?
at 11:01, are the parameters in the formula for a normal distribution flipped?
Ah yes, good catch!
You the best!
What if I don't have an error for each point?
It was discovered that plt.errorbar() does not tolerate any native yerr_data!
Thus, your code needed to be modified as seen here:
yerr_data = np.abs(0.1*np.random.randn(len(x_data)))
YERROR MUST ALWAYS BE POSITIVE TO AVOID TRACEBACK ERRORS!
I Imagine Billy trying to do this. Poor Billy. Not that I would do much better before watching the video 😅
Wait why was i expecting a bird to appear?
We don't want u. We want billy.
It should be N_i(0,1), because you defined N_i(µ, sigma) and always said that sigma should equal 1
otherwise, great video of course
Wait guys ... where is the Birdie ?
hi billy
Where billy
😭 *PromoSM*