def get_random_project_distribution()

in baselines.py [0:0]

• 20 lines of code
• 5 McCabe index (conditional complexity)

    def get_random_project_distribution(self):
        assert os.path.exists("yugongTraces/report-uown-volume-table-20241022-20241028.csv")
        # header: abstractFingerPrint, db_name, table_name, inputDataSize, outputDataSize, cputime
        df = pd.read_csv("yugongTraces/report-uown-volume-table-20241022-20241028.csv")
        prob_in_cloud = self.cloud_target / 100

        # Get unique abstractFingerPrint values
        unique_abFPs = df['abstractFingerPrint'].unique()
        total_cputime = df['cputime'].sum()
        cloud_quota = total_cputime * prob_in_cloud
        onprem_quota = total_cputime * (1 - prob_in_cloud)

        shuffled_abFPs = np.random.permutation(unique_abFPs)

        abFP_status_map = {}
        for abFP in shuffled_abFPs:
            cputime = df[df['abstractFingerPrint'] == abFP]['cputime'].sum()
            if cputime <= cloud_quota:
                abFP_status_map[abFP] = 1
                cloud_quota -= cputime
            else:
                abFP_status_map[abFP] = 0
                onprem_quota -= cputime

        # # Assign Status randomly with probability prob_in_cloud for being 1 (in cloud)
        # abFP_status_map = {abFP: np.random.choice([0, 1], p=[1 - prob_in_cloud, prob_in_cloud]) for abFP in
        #                    unique_abFPs}

        # Map the assigned Status to the dataframe
        df['Status'] = df['abstractFingerPrint'].map(abFP_status_map)

        return df