Link: Probability
When stack data into bins (range), we get Histogram.
We use a curve to approximate the histogram.
Both the histogram and the curve are distibutions, which shows how the probability of measurements are distributed.
Pros:
- It’s more informative, because it’s not limit to discret data points/range.
- It’s more precise, because the curve is also not limited by the width of bins (range).
- For limited data points, use the approximate curve based on mean and sd is usually good enough
Distribution can have many shapes.
- Normal distribution: determined by mean and SD
- Exponential distribution: determined by rate
- Gamma distribution: determined by shape and rate