Course project for the final B.Math career course consisting of a research project about agricultural soil sampling modeling for applying data analysis efficiently to agricultural soils.
the sampled data frame has filtered out rows (stratums) but the real one has all the original records and the merge applies the first row of the real data frame which doesn't match the rows of the sampled data frame. Hence, the relative error is wrong when filtering out stratums.
The sampling API changed to accept floating points percentage as stratum sampling units for each stratum but the simulation still does not work after this update.