Modeling with Common Probability Distributions

1. The Uniform Distribution: All Outcomes Equally Likely

Imagine you're waiting for a shuttle that arrives every 30 minutes. If you arrive at the stop at a random time, your waiting time could be any value between 0 and 30 minutes, and each of those times is equally likely. This is a classic scenario for a Continuous Uniform Distribution. It models a random variable where all outcomes within a specified range [a, b] have an equal chance of occurring.

The PDF is a flat line over the interval [a, b], which is why its graph looks like a rectangle. The height of this rectangle is $\frac{1}{b-a}$ to ensure the total area (total probability) is 1.

To see this in action and learn how to work with it in Python, please watch the first part of this video.

As the video explained, the scipy.stats library is your primary tool for this. Let's summarize the key properties:

• Scenario: Any situation where outcomes in a range $[a,b]$ are equally likely.
• Parameters: a (minimum value, loc in scipy) and b (maximum value, loc + scale in scipy).
• PDF: $f(x)=\frac{1}{b-a}$ for $x\in [a,b]$, and 0 otherwise.
• Expected Value: $E(X)=\mu =\frac{a+b}{2}$ (the midpoint of the interval).
• Variance: $Var(X)={\sigma }^2=\frac{(b-a{)}^2}{12}$.

For more details on the Python implementation, you can refer to the following article.

2. The Exponential Distribution: Waiting Time Between Events

In our last lesson, we used the Poisson distribution to model the number of events in an interval (e.g., 5 customers arriving per hour). The Exponential Distribution is its natural partner: it models the time between those events. It's used to describe the waiting time until the next event in a process where events happen independently and at a constant average rate.

Examples include:

• The time until the next customer calls a support center.
• The time until a radioactive particle decays.
• The lifespan of a component that fails at a constant rate.

A key property of the exponential distribution is that it is memoryless. This means that if a component has already lasted for 100 hours, the probability that it lasts for another 50 hours is the same as the probability that a brand-new component lasts for 50 hours.

Let's watch a short video segment that introduces this distribution.

The distribution is defined by a single parameter, $\lambda$ (lambda), the average rate of events.

• Scenario: Waiting time between events in a Poisson process.
• Parameter: The rate parameter $\lambda$ (e.g., 5 customers/hour).
• PDF: $f(x)=\lambda e^{-\lambda x}$ for $x\ge 0$.
• Expected Value: $E(X)=\mu =\frac{1}{\lambda }$. This is intuitive: if the rate is 5 customers/hour, the average time between customers is 1/5 of an hour.
• Variance: $Var(X)={\sigma }^2=\frac{1}{{\lambda }^2}$.

Important note on Python implementation: In scipy.stats.expon, the scale parameter is used, which is equal to $1/\lambda$. So, if the rate $\lambda$ is 5, you would set scale=1/5.

3. The Normal Distribution: The Bell Curve

The Normal (or Gaussian) Distribution is arguably the most important distribution in all of statistics. It's found everywhere, describing phenomena in both nature and society. This is partly due to the Central Limit Theorem, a powerful idea stating that if you take many random samples and average them, the distribution of those averages will tend to be normal, regardless of the original distribution.

Examples of things that are often normally distributed:

• Heights and weights of a population.
• Measurement errors in an experiment.
• IQ scores.
• Blood pressure.

It is defined by two parameters: the mean $\mu$, which is the center of the distribution, and the standard deviation $\sigma$, which determines its spread.

Standardization and the Z-score

As the video explained, a crucial concept is standardization. To compare different normal distributions (e.g., pizza sizes vs. student heights), we convert them to a Standard Normal Distribution, which has a mean of 0 and a standard deviation of 1. We do this by calculating a Z-score:

$Z=\frac{X-\mu }{\sigma }$

The Z-score tells us exactly how many standard deviations an observation $X$ is from the mean $\mu$.

Working with the Normal Distribution in Python

While Z-tables were traditionally used, Python makes finding these probabilities much easier. Let's see how scipy handles this.

Let's recap the key properties:

• Scenario: Describes many natural phenomena; distribution of sample means.
• Parameters: Mean $\mu$ (loc in scipy) and standard deviation $\sigma$ (scale in scipy).
• PDF: A symmetric, bell-shaped curve centered at $\mu$.
• Expected Value: $E(X)=\mu$.
• Variance: $Var(X)={\sigma }^2$.

Conclusion

Great job! Today you've learned about three essential continuous distributions that are cornerstones of probability and statistics.

Key Takeaways:

Here is a summary table to help you distinguish between them:

The most important skill is recognizing which distribution to apply to a problem. Always think about what the random variable represents: a value from an equally likely range (Uniform), a waiting time (Exponential), or a measurement that clusters around an average (Normal).

Preview of the Next Lesson:

We've now covered the theoretical foundations of the most common probability distributions. In our next lesson, we will get our hands dirty with real data. You will learn how to compute and interpret descriptive statistics (mean, median, variance) for a dataset using Pandas. This will bridge the gap between theoretical probability and practical data analysis.