Pearson correlation in SciPy - Time & Space Complexity

Time Complexity: Pearson correlation

O(n)

Understanding Time Complexity

We want to understand how the time to calculate Pearson correlation changes as the size of the data grows.

How does the number of calculations increase when we have more data points?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np
from scipy.stats import pearsonr

n = 1000  # example size
x = np.random.rand(n)
y = np.random.rand(n)

corr, p_value = pearsonr(x, y)
    

This code calculates the Pearson correlation coefficient between two arrays of length n.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Summation over all elements in the arrays to compute means, variances, and covariance.
• How many times: Each element is visited once in these summations, so n times.

How Execution Grows With Input

As the number of data points n increases, the number of calculations grows roughly in direct proportion.

Input Size (n)	Approx. Operations
10	About 10 operations
100	About 100 operations
1000	About 1000 operations

Pattern observation: Doubling the data size roughly doubles the work needed.

Final Time Complexity

Time Complexity: O(n)

This means the time to compute Pearson correlation grows linearly with the number of data points.

Common Mistake

[X] Wrong: "Calculating Pearson correlation takes quadratic time because it compares every pair of points."

[OK] Correct: The calculation uses sums over the data arrays, not pairwise comparisons, so it only needs to look at each data point once.

Interview Connect

Understanding how Pearson correlation scales helps you explain performance when working with large datasets, a useful skill in data science roles.

Self-Check

"What if we calculated Pearson correlation for multiple pairs of arrays, each of length n? How would the time complexity change?"

Practice

1. What does the Pearson correlation coefficient measure between two variables?

easy

A. The frequency of values in each variable

B. The difference in means between two groups

C. The total sum of values in each variable

D. The strength and direction of a linear relationship

Solution

1. Step 1: Understand Pearson correlation purpose

   Pearson correlation measures how two variables move together in a straight line.

2. Step 2: Identify what it quantifies

   It quantifies both strength (how close points are to a line) and direction (positive or negative slope).

3. Final Answer:

   The strength and direction of a linear relationship -> Option D

4. Quick Check:

   Pearson correlation = strength and direction [OK]

Hint: Pearson = linear relationship strength and direction [OK]

Common Mistakes:

• Confusing correlation with difference in means
• Thinking it measures frequency counts
• Assuming it measures total sums

2. Which of the following is the correct way to import the Pearson correlation function from scipy?

easy

A. from scipy.stats import pearsonr

B. import scipy.pearsonr

C. from scipy import pearsonr

D. import pearsonr from scipy.stats

Solution

1. Step 1: Recall scipy.stats module structure

   The pearsonr function is inside scipy.stats, so import must reflect that.

2. Step 2: Check correct import syntax

   Python import syntax for functions is 'from module import function', so 'from scipy.stats import pearsonr' is correct.

3. Final Answer:

   from scipy.stats import pearsonr -> Option A

4. Quick Check:

   Correct import syntax = from scipy.stats import pearsonr [OK]

Hint: Use 'from scipy.stats import pearsonr' to import [OK]

Common Mistakes:

• Using 'import scipy.pearsonr' which is invalid
• Trying 'from scipy import pearsonr' missing stats
• Wrong syntax like 'import pearsonr from scipy.stats'

3. What is the output of this code?

from scipy.stats import pearsonr
x = [1, 2, 3, 4, 5]
y = [2, 4, 6, 8, 10]
corr, p = pearsonr(x, y)
print("{:.2f}".format(corr))

medium

A. 0.00

B. -1.00

C. 1.00

D. 0.50

Solution

1. Step 1: Analyze the data relationship

   y is exactly 2 times x, so they have a perfect positive linear relationship.

2. Step 2: Understand Pearson correlation output

   Perfect positive linear relationship means correlation coefficient is 1.0.

3. Final Answer:

   1.00 -> Option C

4. Quick Check:

   Perfect positive linear correlation = 1.00 [OK]

Hint: Perfect linear scaling means correlation = 1.00 [OK]

Common Mistakes:

• Confusing correlation with p-value
• Expecting 0.5 for proportional data
• Ignoring perfect linear relationship

4. Identify the error in this code snippet:

from scipy.stats import pearsonr
x = [1, 2, 3]
y = [4, 5]
corr, p = pearsonr(x, y)
print(corr)

medium

A. Lists x and y have different lengths

B. pearsonr function is not imported correctly

C. Missing parentheses in print statement

D. Variables x and y must be numpy arrays

Solution

1. Step 1: Check input data lengths

   pearsonr requires x and y to have the same length. Here, x has 3 elements, y has 2.

2. Step 2: Understand error cause

   Different lengths cause a ValueError when calling pearsonr.

3. Final Answer:

   Lists x and y have different lengths -> Option A

4. Quick Check:

   Input length mismatch causes error [OK]

Hint: Ensure x and y lists have equal length [OK]

Common Mistakes:

• Assuming pearsonr accepts different length inputs
• Thinking print syntax is wrong in Python 3
• Believing inputs must be numpy arrays

5. You have two variables: temperature (in Celsius) and ice cream sales. You want to check if higher temperature relates to higher sales. Which code correctly computes the Pearson correlation and interprets a strong positive relationship?

hard

A. from scipy.stats import pearsonr
corr, p = pearsonr(temp, sales)
if p > 0.7:
print('Strong positive correlation')

B. from scipy.stats import pearsonr
corr, p = pearsonr(temp, sales)
if corr > 0.7:
print('Strong positive correlation')

C. from scipy.stats import pearsonr
corr, p = pearsonr(temp, sales)
if corr < -0.7:
print('Strong positive correlation')

D. from scipy.stats import pearsonr
corr, p = pearsonr(temp, sales)
if abs(corr) < 0.3:
print('Strong positive correlation')

Solution

1. Step 1: Calculate Pearson correlation

   Use pearsonr to get correlation coefficient and p-value between temperature and sales.

2. Step 2: Interpret correlation coefficient

   A strong positive correlation means corr > 0.7. Checking p-value > 0.7 or negative corr is incorrect.

3. Final Answer:

   from scipy.stats import pearsonr corr, p = pearsonr(temp, sales) if corr > 0.7: print('Strong positive correlation') -> Option B

4. Quick Check:

   Strong positive correlation if corr > 0.7 [OK]

Hint: Check corr > 0.7 for strong positive relation [OK]

Common Mistakes:

• Using p-value instead of correlation coefficient
• Checking for negative correlation for positive relation
• Using absolute value less than threshold incorrectly