Example #1
0
def user_uploaded_dataset(request):

    algo_name = request.GET['algo']
    data_name = request.GET['data']

    # get data path and load data
    now = datetime.datetime.now()
    if now.month < 10:
        mth = str(0)+str(now.month)
    else:
        mth = str(now.month)

    if now.day < 10:
        day = str(0)+str(now.day)
    else:
        day = str(now.day)

    path = '/static_media_clustapp/documents/'+str(now.year)+'/'+mth+'/'+day+'/'
    full_path = os.path.realpath('.')
    file_name = full_path+path+data_name
    data_file = scipy.io.loadmat(file_name)

    input_data = data_file['X']
    correct_label = data_file['y']

    # Calling Algorithms
    clustered_label = {}
    unique_label = {}
    if algo_name=="MiniBatchKMeans":
        print 'MiniBatchKMeans'
        kmean_label = algorithm_manager.call_kmean(2, input_data, True)
        clustered_label = kmean_label

    elif algo_name=="Spectral Clustering":
        print "Spectral Clustering"
        spectral_label = algorithm_manager.call_spectral(2, 'arpack', input_data, True)
        clustered_label = spectral_label

    elif algo_name=="Affinity Propagation":
        print "Affinity Propagation"
        affinity_label = algorithm_manager.call_affinity(0.9, input_data, True)
        clustered_label = affinity_label


    # saving the result in the correct directory
    output_data = {}
    output_data ['y'] = clustered_label
    out_file_name = full_path+path+data_name[:-4] + '_result.mat'
    data_file = scipy.io.savemat(out_file_name, output_data)

    # return the path
    return_file_name = path+data_name[:-4]  + '_result.mat'
    return HttpResponse(return_file_name)
Example #2
0
def get_machine_learning_result_given_algorithm_and_dataset(request):
    algo_name = request.GET['algo']
    data_name = request.GET['data']

    print algo_name
    print data_name

    data_dict = {}
    input_data = None
    n_samples = 1500

    # Generateing Sample Data Algorithm
    from sklearn import cluster, datasets 
    if data_name=="Sample Blob":
        blob_dict, X, y = algorithm_manager.generate_blob_sample()
        data_dict = blob_dict
        input_data = X

    elif data_name=="Sample Circle":
        circle_dict,X,y = algorithm_manager.generate_circle_sample()
        #circle_dict,X,y = datasets.make_blobs(n_samples=n_samples, random_state=8)
        data_dict = circle_dict
        input_data = X

    elif data_name=="Sample Moon":
        moon_dict,X,y = algorithm_manager.generate_moon_sample()
        data_dict = moon_dict
        input_data = X

    clustered_label = {}
    unique_label = {}

    # Calling Algorithms
    if algo_name=="MiniBatchKMeans":
        print 'MiniBatchKMeans'
        kmean_label, unique_label = algorithm_manager.call_kmean(2, input_data, False)
        clustered_label = kmean_label

    elif algo_name=="Spectral Clustering":
        print "Spectral Clustering"
        spectral_label, unique_label = algorithm_manager.call_spectral(2, 'arpack', input_data, False)
        clustered_label = spectral_label

    elif algo_name=="Affinity Propagation":
        print "Affinity Propagation"
        affinity_label, unique_label = algorithm_manager.call_affinity(0.9, input_data, False)
        clustered_label = affinity_label

    elif algo_name=="Greedy":
        print "Greedy"
        greedy_label, unique_label = algorithm_manager.call_greedy(2, input_data, False)
        clustered_label = greedy_label

    # main return dictionary contain dictionary about everything
    return_dict = {}
    return_dict['original_data_points'] = data_dict
    return_dict['clustered_data_label'] = clustered_label
    return_dict['clustered_data_unique_label'] = unique_label
    json_return = simplejson.dumps(return_dict)


    return HttpResponse(json_return)