Mathematical Concept of Mean

1. Definition

The mean, also known as the arithmetic average, is a measure of central tendency — a single value that represents the “center” or “typical” value of a dataset.

\[\text{Mean}(\bar{x})= \frac{\sum_{i=1}^{n} x_i}{n}\]

Where:

  • \(x_i\) = individual data points
  • \(n\) = total number of observations

It’s simply the sum of all observations divided by the number of observations.

2. Intuitive Understanding

Think of mean as the balancing point of your data. If each data point were a weight on a scale, the mean is where the scale would balance perfectly.

However, it’s sensitive to extreme values (outliers) — one unusually high or low number can pull the mean significantly in its direction.

3. Example

Dataset: [2, 4, 6, 8, 10]

\[\bar{x} = \frac{2 + 4 + 6 + 8 + 10}{5} = 6\]

So, the mean is 6, representing the center of these values.

4. Types of Mean

There are multiple types of means used in different contexts:

Type Formula Use Case
Arithmetic Mean \(\frac{\sum x_i}{n}\) General average
Weighted Mean \(\frac{\sum w_i x_i}{\sum w_i}\) When values have different importance
Geometric Mean \((\prod x_i)^{1/n}\) For growth rates, ratios
Harmonic Mean \(\frac{n}{\sum \frac{1}{x_i}}\) For rates and ratios (e.g., speed)

For this article, we focus on the arithmetic mean, as it’s most common in statistical analysis.

5. Mathematical Properties

  1. Sum of deviations from the mean is always zero:

    \[\sum (x_i - \bar{x}) = 0\]
  2. Mean minimizes squared deviations:
    The arithmetic mean is the point where the sum of squared deviations is smallest — the foundation of least squares regression.

  3. Linear transformation property:

    \[\text{If } y = a + bx, \text{ then } \bar{y} = a + b\bar{x}\]

NumPy Function for Mean

1. Function: numpy.mean()

Syntax:

numpy.mean(a, axis=None, dtype=None, out=None, keepdims=False)

2. Parameters

Parameter Description
a Input array or object containing numerical data
axis Axis or axes along which to compute the mean. Default is flattened array.
dtype Data type of the returned array.
out Alternative output array.
keepdims If True, retains reduced dimensions as size 1.

3. Example 1: Basic Mean

import numpy as np

data = np.array([2, 4, 6, 8, 10])
mean_value = np.mean(data)
print("Mean:", mean_value)

Output:

Mean: 6.0

4. Example 2: Mean Along an Axis

matrix = np.array([[1, 2, 3],
                   [4, 5, 6],
                   [7, 8, 9]])

mean_row = np.mean(matrix, axis=1)  # Row-wise mean
mean_col = np.mean(matrix, axis=0)  # Column-wise mean

print("Row-wise mean:", mean_row)
print("Column-wise mean:", mean_col)

Output:

Row-wise mean: [2. 5. 8.]
Column-wise mean: [4. 5. 6.]

5. Example 3: Weighted Mean (Manual)

NumPy doesn’t have a built-in weighted mean, but you can compute it easily:

values = np.array([10, 20, 30])
weights = np.array([1, 2, 3])

weighted_mean = np.sum(values * weights) / np.sum(weights)
print("Weighted Mean:", weighted_mean)

Output:

Weighted Mean: 23.333333333333332

6. Numerical Stability Tip

If your data contains very large or very small values, you can use dtype=np.float64 for precision:

np.mean(data, dtype=np.float64)

7. Applications in Data Science

  • Calculating average performance metrics
  • Summarizing sensor or time-series data
  • Finding expected values in probability distributions
  • Normalization and data preprocessing (centering data around zero)

8. Common Mistakes

  • Confusing mean with median (especially in skewed data)
  • Ignoring missing values (use np.nanmean() for datasets with NaN)
  • Always check data distribution before relying solely on mean.

Summary

Concept Formula NumPy Function Sensitive to Outliers Common Use
Mean \(\bar{x} = \frac{\sum x_i}{n}\) np.mean() Yes Average/Expected value