“Data connections” lets you run correlation analysis across different data sources with one click.

Now you’ll have an answer to questions like:

- What’s the connection between my investments and sales?
- What’s the connection between the Facebook posts and people’s perception of my brand?
- Which campaigns seem to work better than others?
- How do different campaign activities affect my business?

How to use Data Connections

Data connections can be used on any graph in a monitor. Create a graph with data from any source, and switch on Data connections under “Edit data series”:

By showing Data connections all relevant series appears in the detail section below the main chart:

What is Data Connection?

Data Connections are based on correlation. All series within a graph is placed into a correlation matrix (available as Advanced option).

Penetrace, the correlation matrix is calculated to show the dependence between two series. The input In of this calculation is a set containing more than two series. The result is a matrix representing the correlation between each pair of these series. With a set of two series, the matrix will be two by two, i.e. four members.

The correlation between two series is calculated according to the following equation:

Assume that we have a set containing three series as follows:

The correlation between the first two series namely {2,3,4} and {2,5,6} is calculated as follows:

And finally:

Note that the value of correlation can be between -1 and 1 where -1 means perfect negative correlation, 0 means no correlation at all and 1 means perfect correlation.

When all nine correlation values are calculated, we can form the resulting correlation matrix which looks like the following table.

Since each series is fully correlated with itself, the main diagonal entries in a correlation matrix are always 1. The correlation matrix is also symmetric as shown above.

## Comments