I am trying to generate some noise to a data set using np.random.randn. Based on the NumPy documentation randn function generates a random number in the standard distribution of mean 0 and variance 1. But standard distribution extends to infinity on both sides, does this mean there is a very very small chance that this function can generate extremely big/small values?
If so how to limit the values with range. I plan to keep 3*sigma threshold on either side. Are there any better ways to generate random normal distribution values within the given range?
