RandomisedDatasetMaster.xlsm

Dataset containing formulas and macro to produce 1000 weighted randomisations of the dataset, which were used in models produced for comparison to "actual" model. The code to import this data to R and produce 1000 models based on these data appear in a text file in this collection.Fields are generally explained by companion dataset NewDataPlay.csvFormulas are fairly complex, but simply assign a random category to each observation (line of data), but weighted so that each individual contributes the same number of observations to each category in each permutation. This removes the bias produced by a single individual contributing many examples to a single category.The macro simply refreshes the dataset (which generates a new permutation), and saves the result, 1000 times.