FREQUENCY TEST USING PYTHON

This is the frequency test equivalent in python code.

Python Dependencies Setup Instructions

Prerequisites

• Ensure that Python is installed on your system.

Installation Steps

Follow these steps to install the necessary Python libraries:

1. Open a terminal or command prompt.
2. Execute the following commands to install the dependencies:

pip install --user numpy
pip install --user pandas
python -m pip install --user scipy

Code Explanation

The provided code snippet defines a function named frequency_test that performs a frequency test on a dataset against a specified distribution. It calculates the expected frequencies based on the chosen distribution, normalizes these frequencies, and then conducts a chi-square test to compare the observed frequencies with the expected frequencies. The function returns the chi-square statistic and the p-value from this test.

Inputs

• data: A NumPy array containing the dataset to be analyzed.
• distribution: A string indicating the type of distribution (e.g., 'uniform', 'binomial', 'poisson', 'exponential', 'geometric') to use for calculating expected frequencies.
• **dist_params: Keyword arguments that provide additional parameters required by the specified distribution.

Process Flow

1. Data Preparation: The input data is flattened, and the unique values along with their counts in the dataset are identified.
2. Expected Frequency Calculation: Depending on the specified distribution, the function calculates the expected frequencies for each unique value in the dataset.
3. Normalization: The expected frequencies are normalized to ensure that their sum matches the sum of the observed frequencies.
4. Chi-Square Test: A chi-square test is performed using the observed frequencies and the normalized expected frequencies.
5. Results: The function returns the chi-square statistic and the p-value resulting from the chi-square test.

Outputs

• chi2_stat: The chi-square statistic, indicating the difference between observed and expected frequencies.
• p_value: The p-value associated with the chi-square statistic, used to determine the statistical significance of the observed difference.

NOTE The dataset provided will only work for a uniform distribution since the value for expected_freq in frequency_test gives a mismatch between the sum of observed frequencies and the sum of expected frequencies; causing a discrepancy beyond a certain tolerance level.

Usage

In the project directory, open a terminal and run python test.py