Skip to main content


Fig. 2 | Environmental Health

Fig. 2

From: Associations between ambient air pollutant mixtures and pediatric asthma emergency department visits in three cities: a classification and regression tree approach

Fig. 2

Regression tree illustrating multipollutant joint effects on pediatric asthma emergency department visits in Atlanta (1999 – 2009). Internal nodes are designated with an oval and numbered such that each node, n, produces two child nodes numbers 2n and 2n + 1. The branches of the tree are labeled according to the level of the pollutant used to partition the tree. For each partition, the branch with the more harmful association is bolded. Terminal nodes are numbered 1A-7A (A for Atlanta). The pie graphs at each terminal node are a graphical representation of the Day-Types that fall into each terminal node. Each pie graph has 12 wedges, four representing each level (L1-L4) of O3 (shades of purple), four representing each level of NO2 (shades of gold), and four representing each level of PM2.5 (shades of blue). Pie wedges are colored if a pollutant level is classified into that terminal node and left white if the pollutant level is absent from the terminal node. Day-Types present in the terminal node can be identified by finding every combination of one O3 wedge (purple), one NO2 wedge (gold) and one PM2.5 wedge (blue). For example terminal node 7A contains 4 Day-Types: O3 level 2, NO2 level 3 and PM2.5 levels 1–4 (2/3/1, 2/3/2, 2/3/3, 2/3/4)

Back to article page