if __name__ == '__main__':
n_cluster = 0
if len(sys.argv) > 1:
n_cluster = int(sys.argv[1])
print 'loading data'
with open('deep_feature.pickle', 'rb') as handle:
NNdict = pickle.load(handle)
with open('pca_feature.pickle', 'rb') as handle:
PCAdict = pickle.load(handle)
#all_data, labeled_data,unlabeled_data,label_unique_list,all_label, labeled_label,all_sample_ID,labeled_sample_ID,unlabeled_sample_ID,gene_names=parse_data.load_integrated_data('data/TPM_mouse_1_4_6_7_8_10_16.txt', whitening=True)
#all_data, labeled_data,unlabeled_data,label_unique_list,all_label, labeled_label, all_weights, labeled_weights, unlabeled_weights,all_sample_ID,labeled_sample_ID,unlabeled_sample_ID,gene_names=parse_data.load_integrated_data('data/TPM_mouse_1_4_6_7_8_10_16.txt',sample_normalize=True,gene_normalize=True)
all_data, labeled_data, unlabeled_data, label_unique_list, all_label, labeled_label, all_weights, labeled_weights, unlabeled_weights, all_sample_ID, labeled_sample_ID, unlabeled_sample_ID, gene_names = parse_data.load_integrated_data(
'data/TPM_mouse_1_4_6_7_8_10_16.txt',
sample_normalize=True,
gene_normalize=True)
#all_data, labeled_data,unlabeled_data,label_unique_list,all_label, labeled_label, all_weights, labeled_weights, unlabeled_weights,all_sample_ID,labeled_sample_ID,unlabeled_sample_ID,gene_names=parse_data.load_integrated_data('data/TPM_mouse_1_4_6_7_8_10_16.txt',whitening=False)
#Label=labeled_label
Label = labeled_label
#bench_k_means(KMeans(init='k-means++', n_clusters=n_digits, n_init=10),
if n_cluster == 0:
n_cluster = max(Label)
print 'n_cluster=', n_cluster
#nlabels=[]
#nlabels.append(max(Label))
n_iter = 10
estimators = []
est_name = []
'''
#data_file_name='../data/TPM_mouse_1_4_6_7_8_10_16.txt'
data_file_name = 'important_file/TPM_mouse_7_8_10_PPITF_gene_9437.txt'
if args.data_file is not None:
data_file_name = args.data_file
#if 'boot' in args.neural_network:
# append_ID='_'+args.neural_network.split('_')[-1].split('.')[0]
#else:
# append_ID=''
if args.identification is not None:
append_ID = '_' + args.identification
else:
append_ID = ''
all_data, labeled_data, unlabeled_data, label_unique_list, all_label, labeled_label, all_weights, labeled_weights, unlabeled_weights, all_sample_ID, labeled_sample_ID, unlabeled_sample_ID, gene_names = parse_data.load_integrated_data(
data_file_name,
sample_normalize=args.sample_normalize,
gene_normalize=args.gene_normalize,
log_trans=args.log_trans,
ref_gene_file=args.reference_gene_file) #'cluster_genes.txt'
all_data_sn1_gn0, _, _, _, _, _, _, _, _, _, _, _, _ = parse_data.load_integrated_data(
data_file_name,
sample_normalize=1,
gene_normalize=0,
ref_gene_file=args.reference_gene_file) #'cluster_genes.txt'
fit_data = all_data
transform_data = all_data
#print 'all_data.shape: ', all_data.shape
#all_data=all_data[:100,:200]
if args.fit == 'labeled':