Matrix Plot Example

Matrix Plot Example

Problem Statement

The following variables are data from three distributions (normal, rayleigh, and uniform). Can you check if the variables are correlated using a matrix plot? 

How to perform analysis

Step 1: Open Sigma Magic
  1. Click on the Sigma Magic button on the Excel toolbar.
  2. Click on the New button to create a new project.
Step 2: Data Preparation
  1. Open the Matrix Plot Plot.xlsx attachment
  2. Select the data and convert it to table format by clicking Ctrl-T.

Step 3: Add analysis template
  1. Click on the Tool Wizard to add the analysis template.
  2. Click on Graph and then Matrix Plot.

Step 4: Specify the analysis options
A new Time Series Plot worksheet will be added to your workbook, and the following dialog box will be opened. Click on the Setup tab on the right. Accept the default values and specify the analysis options.



Next, click on the Data tab and specify the data. Drag and drop the Table1 A, table1 B and Table1 C to the to the Analysis Variables and Table 2 Group to the Categorical Variable.


If you need to change the charts, specify the optional settings in the Charts tab.

 Labels:

  • Add a title for the chart.
  • Label the X-axis and Y-axis appropriately.

Appearance:

  • Adjust colors, font sizes, or other visual elements as needed.
  • Enable/disable gridlines or background shading.


Finally, click the Verify tab to ensure all the inputs are okay and shown in a green checkmark.




Step5: Generate analysis results
Click on the OK button to generate analysis results. The following figure shows an example output screen.

Interpretation of Results

  • The scatterplot matrix visualizes the relationships between variables A, B, and C across different groups.
  • Each subplot shows pairwise relationships, with trend lines fitted to observe possible correlations.
  • The conclusion states that none of the inputs (A, B, C) seem to be correlated with each other.
  • This is evident from the scattered points in the matrix plots, indicating weak or no linear relationships.
  • The two groups (1 & 2) are color-coded to differentiate their behavior within each variable combination.
  • Both groups follow similar trends, meaning there is no significant group-based variation.
  • The table provides N (sample size), Min, Mean, Max, and Standard Deviation (Stdev) for each variable.
  • For example:
    • Variable A (A-1, A-2) has means around 7.81 and 8.46, with slight differences in standard deviations.
    • Variable C (C-1, C-2) has a higher standard deviation (1.85 & 1.83), indicating greater spread/variability.
  • Higher standard deviation values for C (compared to A and B) suggest more variability in C across the dataset.
  • B has the lowest standard deviation, meaning its values are more closely packed around the mean.
  • Since no correlation is detected, predictive modeling using these variables may not be effective without additional data.
  • If looking for influential factors, other variables beyond A, B, and C should be explored to find meaningful patterns.

    • Related Articles

    • Matrix Plot Overview

      A matrix plot is a type of visualization used to display relationships between multiple variables in a dataset. It provides insights into correlations, trends, and patterns by arranging multiple scatter plots or heatmaps in a grid format. Types of ...

    • CE Matrix Example

      Problem Statement Use the C&E Matrix to determine which are the most important causes decline in sales. The team wants to evaluate these causes based on the following criteria: Generation of Sales, Cost, Quality. Use the C&E matrix to determine the ...

    • Scatter Matrix Example

      Problem Statement The data shown in the data tab contains info on adv spend vs. sales revenue. Create a scatter plot between the two variables and draw any conclusions from this exercise. How to perform analysis Step 1: Open Sigma Magic Click on the ...

    • Solution Matrix Example

      Problem Statement Use a solution selection matrix to select the best software. # Notes 1 Software A is the fastest to deploy followed by B & D 2 Software B is the cheapest followed by A & D 3 Software A has medium impact on results (others are lower) ...

    • Overview of Scatter Matrix

      A scatter matrix (also known as a pairplot) is a visualization tool used in statistical analysis and machine learning to understand the relationships between multiple variables in a dataset. It is especially useful for exploratory data analysis (EDA) ...