3. 3-D graphs and more!

We can see two distinct trends in the data. Can we figure out what’s going on? We could try and figure out if industry groups play a role. Let’s add the industry group to Marks -> color.

So it turns out methane is only a contributor for the “Other” industry category, particularly mines (see through facility name when hovering over the data points.)

What about those other measures we have? Can we check those correlations somehow?

3.1. Other contributions to GHGs? (scatterplot matrix)

We can add multiple measures to the Columns shelf to see all the correlations. Let's add AVG biogenic CO2 emissions, AVG Nitrous Oxide emissions, and AVG CO2 emissions (non biogenic)

Paper industry has a lot of “biogenic” CO2 emissions, which are emissions due to biological materials, etc. There is no real trend for N2O. And obviously direct CO2 emissions is a main component of total GHG emissions. Cool.

We could add these same indicators to the rows shelf, which would create a true scatterplot matrix, but we'll skip that in this write-up.

Note that this obscures the number of each facility type. How many paper and pulp facilities are being displayed? Can we get that on this graph somehow? Sure - we need to add another dimension and make this 4-D.

3.2. Create a 4-D scatterplot

We currently have X, Y, and color as dimensions. We’ll need a fourth: Size. Let’s count the number of facilities within each industry group, and make that correspond to the mark size. (This is essentially an aggregation of the previous CH4 chart.)

To do this, we'll use the number of records calculated measure. This is tableau's equivalent of Excel counting everything in a given column, etc. Tableau will count all instances of whatever Dimension is displayed on the graph. In this case, it counts all facilities for each industry group.

From here, we can see that there’s a lot of “Other” facilities and not many “pulp and paper” facilities.

Just for fun, can we add yet another dimension to this?

3.3. 5-D?

We can make this chart more ridiculous by adding our Region Dimensions to Mark Shape. [but never have this many dimensions. The chart is way too complicated at this point] What are all the dimensions here? What is this chart saying?