How Marple is used in science

Marple is a time series visualisation tool. In science, time series data is everywhere. Let's take a look at some water quality data from the Charles river in Boston. The data is publicly available on the website of the US Environmental Protection Agency.

After loading the .csv file, we can make an exploratory visualisation.

An overview of the water quality data

We can very clearly see the correlations between pH, dissolved oxygen, temperature and turbidity. This is because Marple automatically scales all signals to the same range.

Marple automatically scales signals with different value ranges

But the turbidity has some large outliers. It makes the visualisation unclear. The same goes for temperature, it has a false 0 degrees measurement in September. Applying a median filter on both solves the issue:

A clearer overview with a filtered version of turbidity

The water quality standards in the State of Massachusetts demand that water is not to exceed 28.3°C. With a custom function, we can easily verify this.

Marple supports math functions to make calculations

We can see that in the summer months, this maximum temperature has been exceeded a few times.

temp too high

One more interesting correlation to take a look at is between turbidity and phycocyanin. The raw data (top) has a lot of noise, but after applying a low pass filter (bottom) the correlation becomes more clear. Phycocyanin is an indicator for the amount of blue-green algae. The increase in algae might explain the increase in turbidity, or some other variable is at play here.

Marple has filters to clean up your data

A clear correlation between turbidity and phycocyanin


With Marple we were able to

  • Easily load the data
  • See the correlations between signals
  • Clean up the data with median filters and low pass filters
  • Calculate when the temperature was too high