The dataset (p1data.csv) provided in the project folder contains 62 samples collected from colon-cancer patients divided into two classes; there are 22 positive tuples and 40 negative ones. Each tuple (row) consists of the readings for the genes and the class (which is the last column) on one biopsy. Each gene is an attribute. The columns are separated by ",". We number the genes 1 to N in the left-to-right order; we refer to the genes using gi where i is a column number; for example the first gene (column) is called g1.

Your program should work on other datasets with similar formats but they may have different number of rows and different number of columns (perhaps also different class names). [You can assume that there are exactly two classes.] Your program may need a scan of the data to determine the number of genes and the number of instances/rows.

Your (compiled) program will be run using the following command-line command: java P1DM datafilename k m Your program will perform the following two tasks. Task 1. Discretize, rank, and select the top-k genes of the data using the entropy-based method (for 2 intervals). This task produces three files; only the k highest ranked genes (in information gain order) will be included in these files:

1. (a) A file (called geneRankEntropy.txt) containing the entropy based ranking of the k genes, in decreasing information gain order. Each row of the file should contain the gene number, the split value determined by the entropy-based binning method for the gene, and the information gain of the split.