👀 Views: 57 💬 Answers: 1 📅 Created: 2025-06-16

numpy random sampling Python

I'm wondering if anyone has experience with I'm working on a personal project and I'm sure I'm missing something obvious here, but I'm working with an scenario when trying to sample from an array using `np.random.choice` with specified weights. I have an array of elements and corresponding weights, but the output seems to be skewed towards certain values much more than expected. Here's a simplified version of my code: ```python import numpy as np values = np.array([1, 2, 3, 4, 5]) weights = np.array([0.1, 0.1, 0.1, 0.1, 0.6]) # Notice the skewed weights sample_size = 10000 samples = np.random.choice(values, size=sample_size, p=weights) # Count occurrences unique, counts = np.unique(samples, return_counts=True) print(dict(zip(unique, counts))) ``` When I run this, I expect to see most samples being the value `5`, but the counts for values `1`, `2`, and `3` seem disproportionately high. For example, in one run, I got counts like this: ``` {1: 2055, 2: 1982, 3: 2031, 4: 1010, 5: 4922} ``` This doesn't align with the weights I provided. I've tried adjusting the weights slightly, but the scenario continues. I also made sure that the weights sum to 1.0, and I’ve even used the `replace=False` option, but the behavior remains unchanged. Am I misinterpreting how the probabilities work in `np.random.choice`, or is there something else at play? Any insights would be appreciated! I recently upgraded to Python 3.10. Thanks for taking the time to read this! This issue appeared after updating to Python latest. What's the best practice here?