Exploring Gaussian Distribution: Understanding the Math Behind the Bell-Shaped Curve

Exploring Gaussian Distribution Cover

The Gaussian distribution, also known as the normal distribution, is an essential concept in probability theory and statistics that has widespread applications in many fields, such as economics, finance, and the natural sciences. It is named after the famous mathematician Carl Friedrich Gauss, who introduced the concept in the early 19th century.

The Gaussian distribution is defined by its bell-shaped curve, which is symmetric around its mean and characterized by its mean and standard deviation. Understanding the properties and applications of this distribution is crucial for various statistical and machine learning techniques, making it a fundamental concept in data science.

In this blog post, we will delve deeper into the Gaussian distribution and explore its mathematical definition, properties, and numerous applications. By the end of this post, you will have a solid understanding of this important concept and be equipped to use it in various data analysis tasks.

Probability Density Function (PDF)

The Gaussian distribution is characterized by its probability density function (PDF), which is given by the following formula:

f(x)=1σ2πe(xμ)22σ2\begin{aligned} f(x) = \frac{1}{\sigma \sqrt{2 \pi}} e^{-\frac{(x-\mu)^2}{2 \sigma^2}} \end{aligned}

Here, μ\mu is the mean of the distribution, and σ\sigma is its standard deviation. The formula tells us that the probability density of a value xx in the distribution is proportional to the distance of xx from the mean, and inversely proportional to the spread of the distribution. In other words, the probability of finding a value xx in a given range is higher if that range is closer to the mean and narrower if the standard deviation is small.

The Empirical Rule and the Z-Score

The empirical rule is a statistical rule of thumb that can be used to estimate the proportion of values within a certain number of standard deviations from the mean in a Gaussian distribution. It is also known as the 68-95-99.7 rule or the three-sigma rule.

According to the empirical rule:

  • About 68% of the values in a Gaussian distribution will fall within one standard deviation of the mean.
  • About 95% of the values will fall within two standard deviations of the mean.
  • About 99.7% of the values will fall within three standard deviations of the mean.

The following image shows a Gaussian distribution curve with the empirical rule overlaid:

Empirical Rule

The green area in the image represents the proportion of values that fall within one standard deviation of the mean, the yellow area represents the proportion of values that fall within two standard deviations of the mean, and the red area represents the proportion of values that fall within three standard deviations of the mean.

To measure how far a value is from the mean in a Gaussian distribution, we can use the z-score, which is a measure of how many standard deviations a value is from the mean. The formula for calculating the z-score is:

z=xμσ\begin{aligned} z = \frac{x - \mu}{\sigma} \end{aligned}

Here, xx is the value being measured, μ\mu is the mean of the distribution, and σ\sigma is the standard deviation of the distribution. The z-score indicates how many standard deviations a value is above or below the mean of the distribution. For example, if the z-score of a value xx is 1, it means that xx is one standard deviation above the mean. If the z-score is -1, it means that xx is one standard deviation below the mean.

The area under the curve of the Gaussian distribution represents the probability of a value falling within a certain range. The total area under the curve is equal to 1, which means that the sum of the probabilities of all possible outcomes is 1. The area under the curve between two z-scores represents the probability of a value falling between those two values. For instance, the area under the curve between a z-score of -1 and a z-score of 1 represents the probability of a value falling within one standard deviation of the mean, which is approximately 68%. We can use the z-table to find the area under the curve between two z-scores.

Solving an Example Problem

Let's apply the concepts of the empirical rule and z-score to solve an example problem.

Problem Statement:

Suppose that the weights of a certain population of people follow a Gaussian distribution with a mean of 70 kg and a standard deviation of 5 kg. What is the probability that a randomly selected person from this population weighs between 60 kg and 80 kg?

Solution:

First, we can calculate the z-scores for the weights of 6060 kg and 8080 kg using the formula:

z60=60705=2z_{60} = \frac{60 - 70}{5} = -2 z80=80705=2z_{80} = \frac{80 - 70}{5} = 2

Next, we can use a z-value table to find the area under the curve between the z-scores of 2-2 and 22. From the table, we find that the area under the curve between a z-score of 2-2 and 22 is approximately 0.95450.9545.

Therefore, the probability that a randomly selected person from this population weighs between 6060 kg and 8080 kg is approximately 0.95450.9545, or 95.45%95.45\%.

Using the empirical rule, we can also estimate the proportion of weights that fall within one, two, and three standard deviations from the mean. Since the standard deviation is 55 kg, we can calculate the weight ranges as follows:

  • One standard deviation: 7070 kg ±\pm 1×51 \times 5 kg == [65[65 kg, 7575 kg]]
  • Two standard deviations: 7070 kg ±\pm 2×52 \times 5 kg == [60[60 kg, 8080 kg]]
  • Three standard deviations: 7070 kg ±\pm 3×53 \times 5 kg == [55[55 kg, 8585 kg]]

From the ranges, we can see that the weight range of 6060 kg to 8080 kg falls within two standard deviations from the mean. Therefore, we would expect approximately 95%95\% of the population to have weights within this range, which agrees with the probability we calculated using z-scores and the z-value table.

Creating a Gaussian Distribution Plot in Python

A Gaussian distribution, also known as a normal distribution, is a probability distribution that is commonly found in nature and statistics. In Python, we can easily generate a Gaussian distribution using the NumPy library and visualize it using Matplotlib.

Here is a step by step guide to creating a Gaussian distribution plot in Python:

1

Import the required libraries

The first step is to import the necessary libraries, which are numpynumpy and matplotlibmatplotlib:

gaussian_distribution.py
import numpy as np
import matplotlib.pyplot as plt
gaussian_distribution.py
import numpy as np
import matplotlib.pyplot as plt
2

Generate random data

Next, we can generate some random data from a Gaussian distribution using the numpy.random.normal()numpy.random.normal() function. This function takes three arguments: the mean, the standard deviation, and the number of data points to generate. For example, to generate 1000 data points with a mean of 0 and a standard deviation of 1, we can use the following code:

gaussian_distribution.py
data = np.random.normal(0, 1, 1000)
gaussian_distribution.py
data = np.random.normal(0, 1, 1000)
3

Create a histogram

We can create a histogram of the data using the matplotlib.pyplot.hist()matplotlib.pyplot.hist() function. This function takes the data as input, as well as some additional optional arguments, such as the number of bins to use for the histogram. Here's an example of how to create a histogram with 20 bins:

gaussian_distribution.py
plt.hist(data, bins=20, density=True, alpha=0.6)
gaussian_distribution.py
plt.hist(data, bins=20, density=True, alpha=0.6)

In this code, the density=Truedensity=True argument means that the histogram will be normalized to represent a probability distribution, and the alpha=0.6alpha=0.6 argument means that the bars in the histogram will be slightly transparent.

4

Create a Gaussian distribution plot

To plot a smooth bell curve that fits the histogram, we can use the numpynumpy and matplotlibmatplotlib libraries to calculate and plot a probability density function (PDF) for the Gaussian distribution. Here's an example of how to do this:

gaussian_distribution.py
# Calculate the mean and standard deviation of the data
mu, std = np.mean(data), np.std(data)
 
# Create an array of x-values ranging from mu - 4*std to mu + 4*std
x = np.linspace(mu - 4*std, mu + 4*std, 100)
 
# Calculate the PDF for the Gaussian distribution using the x-values, mean, and standard deviation
pdf = 1/(std * np.sqrt(2 * np.pi)) * np.exp(- (x - mu)**2 / (2 * std**2))
 
# Plot the PDF as a line on top of the histogram
plt.plot(x, pdf, color='red')
gaussian_distribution.py
# Calculate the mean and standard deviation of the data
mu, std = np.mean(data), np.std(data)
 
# Create an array of x-values ranging from mu - 4*std to mu + 4*std
x = np.linspace(mu - 4*std, mu + 4*std, 100)
 
# Calculate the PDF for the Gaussian distribution using the x-values, mean, and standard deviation
pdf = 1/(std * np.sqrt(2 * np.pi)) * np.exp(- (x - mu)**2 / (2 * std**2))
 
# Plot the PDF as a line on top of the histogram
plt.plot(x, pdf, color='red')

In this code, the np.linspace()np.linspace() function creates an array of x-values ranging from 4 standard deviations below the mean to 4 standard deviations above the mean, and the pdfpdf variable is calculated using the formula for the Gaussian probability density function. Finally, the plt.plot()plt.plot() function is used to plot the PDF as a red line on top of the histogram.

5

Add labels and show the plot

Finally, we can add labels to the plot and show it using the matplotlib.pyplot.xlabel()matplotlib.pyplot.xlabel(), matplotlib.pyplot.ylabel()matplotlib.pyplot.ylabel(), and matplotlib.pyplot.show()matplotlib.pyplot.show() functions:

gaussian_distribution.py
plt.xlabel('Data')
plt.ylabel('Probability Density')
plt.title('Histogram of Random Data with Fitted Gaussian Distribution Curve')
plt.show()
gaussian_distribution.py
plt.xlabel('Data')
plt.ylabel('Probability Density')
plt.title('Histogram of Random Data with Fitted Gaussian Distribution Curve')
plt.show()

Putting it all together, the complete code to generate a Gaussian distribution plot in Python looks like this:

gaussian_distribution.py
import numpy as np
import matplotlib.pyplot as plt
 
# Generate random data from a Gaussian distribution
data = np.random.normal(0, 1, 1000)
 
# Create a histogram of the data
plt.hist(data, bins=20, density=True, alpha=0.6)
 
# Calculate the mean and standard deviation of the data
mu, std = np.mean(data), np.std(data)
 
# Create an array of x-values ranging from mu - 4*std to mu + 4*std
x = np.linspace(mu - 4*std, mu + 4*std, 100)
 
# Calculate the PDF for the Gaussian distribution using the x-values, mean, and standard deviation
pdf = 1/(std * np.sqrt(2 * np.pi)) * np.exp(- (x - mu)**2 / (2 * std**2))
 
# Plot the PDF as a line on top of the histogram
plt.plot(x, pdf, color='red')
 
# Add labels and show the plot
plt.xlabel('Data')
plt.ylabel('Probability Density')
plt.title('Histogram of Random Data with Fitted Gaussian Curve')
plt.show()
gaussian_distribution.py
import numpy as np
import matplotlib.pyplot as plt
 
# Generate random data from a Gaussian distribution
data = np.random.normal(0, 1, 1000)
 
# Create a histogram of the data
plt.hist(data, bins=20, density=True, alpha=0.6)
 
# Calculate the mean and standard deviation of the data
mu, std = np.mean(data), np.std(data)
 
# Create an array of x-values ranging from mu - 4*std to mu + 4*std
x = np.linspace(mu - 4*std, mu + 4*std, 100)
 
# Calculate the PDF for the Gaussian distribution using the x-values, mean, and standard deviation
pdf = 1/(std * np.sqrt(2 * np.pi)) * np.exp(- (x - mu)**2 / (2 * std**2))
 
# Plot the PDF as a line on top of the histogram
plt.plot(x, pdf, color='red')
 
# Add labels and show the plot
plt.xlabel('Data')
plt.ylabel('Probability Density')
plt.title('Histogram of Random Data with Fitted Gaussian Curve')
plt.show()

When you run this code, it will generate a plot that shows the histogram of the random sample, along with a smooth bell-shaped curve that represents the probability density function of the Gaussian distribution:

Fitted Gaussian Distribution

This graph is a useful visual representation of a Gaussian distribution, and can be used to better understand the properties of this type of distribution.

Properties of the Gaussian Distribution

The Gaussian distribution has several important properties:

  • Symmetry: The Gaussian distribution is symmetric around its mean, which means that the left and right sides of the curve are mirror images of each other.
  • Unimodality: The Gaussian distribution has only one peak, so it is referred to as a unimodal distribution.
  • Asymptotic: The Gaussian distribution approaches 00 as xx approaches infinity or negative infinity.
  • Normalization: The Gaussian distribution integrates to 11 over the entire real line, which means that the total area under the curve is equal to 11.
  • Characterized by mean and standard deviation: The Gaussian distribution is completely characterized by its mean and standard deviation.

Using Gaussian Distribution in Real-World Applications

The Gaussian Distribution is used in many real-world applications, including:

  • Image Processing: Gaussian Distribution is used in image processing to filter out noise from images.
  • Data Analysis: Gaussian Distribution is used in data analysis to describe the distribution of data and to make predictions about future data.
  • Machine Learning: Gaussian Distribution is used in machine learning algorithms, such as Gaussian Naive Bayes and Gaussian Mixture Models, to model the distribution of data and make predictions.
  • Weather Prediction: Gaussian Distribution is used in weather prediction to model the distribution of temperature and precipitation in a given region.

Conclusion

In this article, we learned about the Gaussian Distribution, which is a continuous probability distribution that is symmetric, unimodal, and characterized by its mean and standard deviation. We also learned how to generate a Gaussian Distribution plot in Python using the numpynumpy and matplotlibmatplotlib libraries, and we learned about some of the real-world applications of the Gaussian Distribution. Thank you for reading!



Mir Sazzat Hossain is a Research Assistant at the Independent University of Bangladesh's Center for Computation and Data Science (CCDS).

Comments

Do you have a problem, want to share feedback, or discuss further ideas? Feel free to leave a comment here! Please stick to English. This comment thread directly maps to a discussion on GitHub, so you can also comment there if you prefer.

Instead of authenticating the giscus application, you can also comment directly on GitHub.

Related Articles

The Poisson Distribution: Your Key to Predicting the Unforeseeable

The blog post explores the Poisson distribution, a statistical distribution commonly used in various fields, including astronomy, to model random events. It explains the properties of the distribution, its real-world applications, and provides a step-by-step guide on how to visualize the Poisson distribution using Python. The post also discusses the issue of Poisson noise in astronomical observations and presents a practical example of how to calculate and deal with it using Python.