I'm building a generic macro to put into SAS to run every time I get a new data set. It tests for a lot of things........mean,gini, number of unique values, variance, number of negative values.....a lot of stats come out -- about 50+
I know there can't be a perfect catch all, but I"m curious what you might look for?
Remember this is just as 'practical' as it is theoretical. For example the negative values is huge...if I get data that has money spent in a certain area, i wouldn't expect to see a negative value. But this happens a lot.
Again, be curious to see what you guys come up with?
I know there can't be a perfect catch all, but I"m curious what you might look for?
Remember this is just as 'practical' as it is theoretical. For example the negative values is huge...if I get data that has money spent in a certain area, i wouldn't expect to see a negative value. But this happens a lot.
Again, be curious to see what you guys come up with?