What we can see from Fig. 5 is that a cluster of low revenue district is seen in the center of the municipality. Districts having high median revenue are dispatched all over the geographical space. However, we see a trend where in the north-west of the municipality as well as in the extreme south-east, districts have a quite high income. Fig. 6 shows the relationship between the median revenue and the proportion of pollutes sites. Note that this scatter plot included the districts having no data. The trend of this data representation is shown in purple. However, we thought it would be more relevant to exclude those polygon that is why, thanks to Geoda tools, we represented on the same graph the same data set but excluding the three districts having any median revenue, and the corresponding the trend line is shown in red. As a general remark, we see there is a strong correlation between the median revenue and the proportion of polluted sites : as the proportion of polluted sites increases in the district, we have good chance to have a lower median revenue in this district.