Correlation Detective Logo

Correlation Detective

Correlation Detective is the state-of-the-art algorithm for mining multivariate correlations in data. In this demo, you can run Correlation Detective in your browser, on your own data. For more information on multivariate correlations and their applications, see the website.

Note: This demo runs entirely in your browser, no data is sent to any server. However, depending on the size of your data, it may use a lot of memory and CPU. If your browser crashes, try using a smaller dataset, or limiting the number of dimensions and/or vectors to read below. To run Correlation Detective on larger datasets, consider running it in Java, see the GitHub repository.

📂 1. Load Data

Load your own CSV file to detect correlations, or use one of our sample datasets.

Here you can configure how the input is loaded, and discard parts of the dataset.

The delimiter that separates the fields in the CSV file.

The maximum amount of vectors to read. Any extra vectors are ignored. (When empty, this is infinity)

The maximum amount of events to read. Any other events are ignored. (When empty, this is infinity)

Format:
More options

The partition to read, when this is nonzero, the first partition * maxDimensions events are ignored.

👀 Preview

⚙️ 2. Define Query

When data has been loaded, you can run CorrelationDetective over it.

Query Type:
Pattern:

The maximum amount of variables on the left hand side (LHS) and right hand side (RHS) to find a correlation between.

More options
Additional constraints:

📊 Results

✨ 3. Analyze