unfold_dict[label].extend([v] * n)
# clean values (try to convert to float and remove leading ;)
for k in unfold_dict:
unfold_dict[k] = [_clean_value(v) for v in unfold_dict[k]]
return pd.DataFrame(unfold_dict, columns=df.columns)
###################################################################
if __name__ == "__main__":
from metabolinks.datasets import demo_dataset
# read sample data set demo2
print('\nLoad demo2: data with labels ------------\n')
demo2 = demo_dataset('demo2')
data = demo2.data
y = demo2.target
print('-- info --------------')
print(data.transpose().cdl.info())
print('-- global info---------')
print(data.transpose().cdl.info(all_data=True))
print('-----------------------')
print(data)
print('\n--- fillna_zero ----------')
new_data = fillna_zero(data)
print(new_data)
print('--- fillna_value 10 ----------')
new_data = fillna_value(data, value=10)
print(new_data)
right_index=True)
return annotations
if __name__ == '__main__':
from metabolinks import datasets
dbs = load_local_dbs()
# for k, v in dbs._asdict().items():
# print(f'----- {k}')
# print(v)
# print('*'*30)
df = datasets.demo_dataset('masstrix_output').data
print("Demo data:\n")
df.info()
print('************************')
print(df.KEGG_cid)
print('************************')
# get one identifier per row
cids = df.KEGG_cid.str.split('#').explode()
print('\n\n----- Identifier list')
print(cids)
print('-----------------------------------')
# build identifier translation table
identifiers = get_identifiers(cids, dbs, trace=False)
else:
comps.append('other')
return comps
if __name__ == '__main__':
from io import StringIO
from metabolinks import datasets
from metabolinks.dataio import read_MassTRIX
print('------ test element_composition() ------')
for test in 'C11H24NO7P', 'C13H19ClN2O2', 'C12H21O11R':
print(test, '->', element_composition(test))
print('\n------ test insert_element_counts() ------')
df = datasets.demo_dataset('table_with_formulae').data
print(df)
print('+++++ after insertion ++++++')
dfi = insert_element_counts(df)
print(dfi)
print('\n------ test element_composition_series ------')
# file_name = "MassTRIX_output.tsv"
# import os
# _THIS_DIR, _ = os.path.split(os.path.abspath(__file__))
# testfile_name = os.path.join(_THIS_DIR, "data", file_name)
df = read_MassTRIX(
StringIO(datasets.create_demo('masstrix_output').as_str()))
def cleanup_cols(df, isotopes=True, uniqueID=True, columns=None):
# empty subset, skip
continue
else:
# common features with exactely len(t) ocorrences
if len(t) == 1:
all_feats = objects[t[0]]
else:
all_feats = common([objects[i] for i in t])
features = all_feats[all_feats.isin(count_groups[len(t)-1])]
subset_names = tuple([names[i] for i in t])
res[subset_names] = features
return res
if __name__ == "__main__":
print('Demo data with labels------------\n')
dataset = datasets.demo_dataset('demo2').data.transpose()
print(dataset)
print('-- info --------------')
print(dataset.cdl.info())
print('-- global info---------')
print(dataset.cdl.info(all_data=True))
print('\n***** SIMILARITY MEASURES ****')
similarities = mz_similarity(dataset, has_labels=True)
print(similarities)
print('\n\n***** FEATURE OVERLAP (AND VENN DIAGRAM CALCULATIONS ****')
print('--- example data sets')
s1 = pd.DataFrame({'Bucket label': ['A0', 'A1', 'A2', 'A3'],
'Name': ['B0', np.nan, 'B2', 'B3'],
'Formula': ['C0', 'C1', 'C2', 'C3']},
index=[0, 1, 2, 3]).set_index('Bucket label')
def load_demo2():
return demo_dataset('demo2')