def run_msms_data(fragment, neutral_loss, mzdiff, ms1, ms2):
if len(sys.argv) > 1:
K = int(sys.argv[1])
else:
K = 300
print "Cross-validation for K=" + str(K)
n_folds = 4
n_samples = 500
n_burn = 250
n_thin = 5
alpha = 50.0 / K
beta = 0.1
is_num_samples = 10000
is_iters = 1000
ms2lda = Ms2Lda.lcms_data_from_R(fragment, neutral_loss, mzdiff, ms1, ms2)
df = ms2lda.df
vocab = ms2lda.vocab
cv = CrossValidatorLda(df, vocab, K, alpha, beta)
cv.cross_validate(n_folds,
n_burn,
n_samples,
n_thin,
is_num_samples,
is_iters,
method="with_mixture")
def run_msms_data(fragment, neutral_loss, mzdiff,
ms1, ms2):
if len(sys.argv)>1:
K = int(sys.argv[1])
else:
K = 300
print "Cross-validation for K=" + str(K)
n_folds = 4
n_samples = 500
n_burn = 250
n_thin = 5
alpha = 50.0/K
beta = 0.1
is_num_samples = 10000
is_iters = 1000
ms2lda = Ms2Lda.lcms_data_from_R(fragment, neutral_loss, mzdiff, ms1, ms2)
df = ms2lda.df
vocab = ms2lda.vocab
cv = CrossValidatorLda(df, vocab, K, alpha, beta)
cv.cross_validate(n_folds, n_burn, n_samples, n_thin,
is_num_samples, is_iters, method="with_mixture")
from lda_for_fragments import Ms2Lda
ms2lda = Ms2Lda.resume_from('notebooks/results/beer3pos.project')
ms2lda.do_thresholding(th_doc_topic=0.05, th_topic_word=0.01)
ms2lda.print_topic_words()
special_nodes = [
('doc_21213', '#CC0000'), # maroon
('doc_21758', 'gold'),
('doc_21182', 'green'),
('topic_240', '#CC0000'), # maroon
('topic_76', 'aqua'),
('topic_253', '#ff1493') # deep pink
]
ms2lda.plot_lda_fragments(consistency=0.0, interactive=True, to_highlight=special_nodes)