Refactor loadFromCSV to include k-fold code

[kallewesterling]

See BNdata.loadFromCSV: should rewrite the loadFromCSV function as loaddataset_kfold and write the kfold code in here and return list of lists of indexes

def loadFromCSV (self, header=False):
    # TODO: should rewrite the loadFromCSV function as loaddataset_kfold and write the kfold code in here and return list of lists of indexes # Issue #7 --> extract to separate issue
    dataset = []
    with open(self.file, 'rb') as csvfile:
        lines = csv.reader(csvfile)

        for row in lines:
            dataset.append(row)
    data = []
    if (header == True): data.append(dataset[0])
    for i in range(0, len(dataset)):
        row = []
        for j in range(0, len(dataset[i])):
            if i == 0:
                row.append(dataset[i][j])
            else:
                item = float(dataset[i][j])
                row.append(item)
        data.append(row)
    self.data = data

    return data