I have a dataset that has around 4 million cases and a lot of variables, my dependant variables are continuous values representing either gas or electricity consumption and the independent variables are all categorical representing a range of indexes and household attributes.
The problem that I am having is that I want to run a one-way independent ANOVA but because my data violates the homogeneity of variance assumption I am planning on using the Kruskal-Wallis Test instead. When I run the K-S test I get the following warning:
"There is insufficient workspace memory to process all the cases. Break up the request, rerun with more workspace, or use the SAMPLE subcommand. You can increase workspace with the SET WORKSPACE command. Execution of this command stops."
I have tried increasing the Workspace memory through the syntax window and have increased it up to 2GB but I still get the same result. Even when tried to test a sample of my data it would only run the test on a 1% sample. Obviously I want to run the test on all valid cases. Is there a way around this issue as I am stuck on what else I can do?
Any help would be much appreciated.