There is already an answer of this question HERE!
Completing the answer:
Basically you need to link the plot to an area so that the plot expansion factor can be calculated.
If you have an attribute in your plot that is called region (for instance), then you can tell Collect Earth how big the regions are. Then when you run Saiku, Collect Earth will calculate the expansion factor by dividing the are of the region by the number of plots that have the attribute region set to that one.
So, in case the regions in your survey are : Los Rios and Atacama you would add an areas_per_attribute.csv file to your survey (using the files feature in the main tab of the Survey Designer) with the following data:
region,area,weight
Los Rios,1838148,1
Atacama ,7564284,1
IMPORTANT : the name of the columns are very important. Both area (in hectares) and weight should be the last two columns. Then the name(header) of the first column must coincide with the name of the attribute that you want to link from the survey (in this case we would have a Text Attribute called region in the survey )
Weight
Weight is usually set to 1. But, if you had a sampling design where there are different intensities by region this can be handy as it allows to perform calculation in Saiku by weighted numbers of plots. E.g. in "Los rios" as it is a forested area you have one plot every 2000 meters, but in "Atacama" as it is a desertic area you will have one plot every 8000 meters. Thus the weight of the points in Atacama is higher than those in Los Rios. Then you will have something like this:
region,area,weight
Los Rios,1838148,1
Atacama ,7564284,16
In this case, the weight of the plots of Atacama is 16, as there are 16 times LESS plots in Atacama compared to Los Rios (if they had the same area)
