This are the python codes used for the paper "Recomputing the Hubble constant", which is written as part of the Youth and Science program by Fundació Catalunya La Pedrera.
In summer 2024, I discovered np.polyfit prioritizes minimizing the vertical distance to data points, which can lead to wrong/unprecise values. Using scipy.ord module avoids this error, as it accounts equally for both axis (x and y directions), thus usually preferred for scientific curve fitting. This is something I wasn't aware when I conducted this project in 2023 (11th grade), as this was the first time I was using python for research and my first big project. I'll use scipy from now on ;)
Using scipy instead of numpy my data results in a value for H0 of numpy it results in a similar value, confirming the axis-bias when using polyfit.
In March 2025 I update both the code and the research paper in this repository. Note that, altough two years later I found some things that could be enhanced, I only corrected the values and lines refering to the here mentioned issue.