DAMACRO

This repository presents the DAMACRO framework which utilize pre-trained classifiers to clustering the data into different cluster for compression ratio improvment

Code usage

Run the scripts by using following command

python main.py

Parameter setting

Here is some parameters that can be changed to get experiment results.

    original_data_sizes=[50.12,40.09,45.11] 
    scaler_data_size = 71.78
    file_path = 'data/original'
    dataset = 'flight_adjusted' 

    train_percents = [0.8,0.2,0.1] 
    model_path = 'models'
    model_names = ['DecisionTree','GaussianNB','LogisticRegression','MLP','QDA']
    alg = ['gzip','lz4','zstd']
    overlaps = ['insert','no_overlap','update']
    cat_i = [0, 1, 3, 6]#[0, 5]#[0, 1, 2, 3] #[0, 5]#[0, 5]#[0, 1, 3, 6]
    num_i = [2, 5]#[3, 6]#[3, 6]#[3, 6]#[2, 5]
    delimiter = '|'
    chunk_size = 10000

    target_path = '' #put target path here
    networkspeeds =[25,20,15,10,5]#,20,15,10,5,2,20,15
    target_ip = '' #put target ip here
    target_user = '' #put user name here