def post(self, request):
serializer = IncomingDataSerializer(data=request.data)
if serializer.is_valid():
phrase_data = assnake.Dataset(serializer.data)
print(phrase_data)
return JsonResponse(data={'status': phrase_data})
return Response(serializer.errors, status=status.HTTP_400_BAD_REQUEST)
def prepare_sample_set_tsv_and_get_results(sample_set_dir_wc, result_wc, df,
sample_sets, overwrite, **kwargs):
res_list = []
destroy_if_not_run = {'directories': [], 'files': []}
df_loaded = assnake.Dataset(df)
for sample_set_name in sample_sets.keys():
sample_set_dir = sample_set_dir_wc.format(
fs_prefix=df_loaded.fs_prefix, df=df, sample_set=sample_set_name)
sample_set_loc = os.path.join(sample_set_dir, 'sample_set.tsv')
sample_set = sample_sets[sample_set_name]
if not os.path.exists(sample_set_dir):
os.makedirs(sample_set_dir, exist_ok=True)
destroy_if_not_run['directories'].append(sample_set_dir)
if not os.path.isfile(sample_set_loc):
sample_set.to_csv(sample_set_loc, sep='\t', index=False)
destroy_if_not_run['files'].append(sample_set_loc)
else:
click.secho('Sample set with this name already exists!')
if overwrite:
sample_set.to_csv(sample_set_loc, sep='\t', index=False)
click.secho('Overwritten')
res_list += [
result_wc.format(fs_prefix=df_loaded.fs_prefix,
df=df_loaded.df,
sample_set=sample_set_name,
**kwargs)
]
return res_list, destroy_if_not_run
def update_fs_samples_csv(dataset):
'''
Scans dataset folder and updates fs_samples.tsv if necessary
:param dataset: Name of the dataset
:return: Returns sample dict in loc
'''
fs_samples_tsv_loc = '{assnake_db}/datasets/{df}/assnake_samples.tsv'.format(assnake_db=read_assnake_instance_config()['assnake_db'], df=dataset)
df = assnake.Dataset(dataset)
fs_samples_pd = pd.concat(list(df.sample_sets.values()))
fs_samples_pd = df.sample_sets['raw']
fs_samples_pd['final_preprocessing'] = 'never_set'
fs_samples_pd.to_csv(fs_samples_tsv_loc, sep='\t', index = False)
return True
def generic_command_dict_of_sample_sets(config, df, preproc, meta_column,
column_value, samples_to_add,
exclude_samples, **kwargs):
'''
This returns several sample sets.
'''
df_loaded = assnake.Dataset(df)
sample_sets_dict = {}
meta_loc = os.path.join(df_loaded.full_path, 'df_samples.tsv')
if os.path.isfile(meta_loc):
meta = pd.read_csv(meta_loc, sep='\t', index_col=0)
if meta_column is not None:
if column_value is not None:
sample_set, sample_set_name = generic_command_individual_samples(
config, df, preproc, meta_column, column_value,
samples_to_add, exclude_samples, **kwargs)
sample_sets_dict.update({sample_set_name: sample_set})
else: # treat empty column_value as creating multiple sample_sets for each column_value
column_values = list(meta[meta_column].unique())
for column_value in column_values:
sample_set, sample_set_name = generic_command_individual_samples(
config, df, preproc, meta_column, column_value,
samples_to_add, exclude_samples, **kwargs)
sample_sets_dict.update({sample_set_name: sample_set})
else:
sample_set, sample_set_name = generic_command_individual_samples(
config, df, preproc, meta_column, column_value, samples_to_add,
exclude_samples, **kwargs)
sample_sets_dict.update({sample_set_name: sample_set})
else:
if meta_column is not None:
click.echo(
'A metadata column is specified, but there is no metadata file: %s'
% meta_loc,
fg='red')
exit()
sample_set, sample_set_name = generic_command_individual_samples(
config, df, preproc, meta_column, column_value, samples_to_add,
exclude_samples, **kwargs)
sample_sets_dict.update({sample_set_name: sample_set})
return sample_sets_dict
def generic_command_individual_samples(config, df, preproc, meta_column,
column_value, samples_to_add,
exclude_samples, **kwargs):
"""
Construct sample sets, has multiple options.
Returns dict or sample_sets based on the provided options.
meta_column - factor column in sample metadata sheet. Cannot be combined with samples_to_add. exclude_samples has higher proirity.
column_value - value of column to select by. \
Can be multiple - separated by commas without whitespace. \
If meta_column is provided, bot no column_value is provided \
- treat it like select all unique values of that column.
If multiple - one value - one sample_set. If --merge enabled - all values go in one sample_set
assnake result request megahit -d FMT_FHM -c source run
"""
exclude_samples = [] if exclude_samples == '' else [
c.strip() for c in exclude_samples.split(',')
]
samples_to_add = [] if samples_to_add == '' else [
c.strip() for c in samples_to_add.split(',')
]
df_loaded = assnake.Dataset(df)
config['requested_dfs'] += [df_loaded.df]
# Now for the meta column stuff
meta_loc = os.path.join(df_loaded.full_path, 'df_samples.tsv')
if os.path.isfile(meta_loc):
meta = pd.read_csv(meta_loc, sep='\t', index_col=0)
if meta_column is not None:
if column_value is not None:
subset_by_col_value = meta.loc[meta[meta_column] ==
column_value]
if len(subset_by_col_value) > 0:
samples_to_add = list(subset_by_col_value.index.values)
samples_to_add = list(
set(df_loaded.sample_containers['df_sample'].values).
intersection(set(samples_to_add)))
if len(samples_to_add) == 0:
click.secho('There are 0 samples for %s == %s' %
(meta_column, column_value),
fg='red')
exit()
if preproc is None:
# LONGEST
preproc = max(list(df_loaded.sample_sets.keys()), key=len)
click.echo(
'Preprocessing is not specified, using longest for now - %s' %
preproc)
sample_set = assnake.api.loaders.load_sample_set(
config['wc_config'],
df_loaded.fs_prefix,
df_loaded.df,
preproc,
samples_to_add=samples_to_add)
if len(exclude_samples) > 0:
sample_set = sample_set.loc[
~sample_set['df_sample'].isin(exclude_samples), ]
# click.echo(tabulate(sample_set[['df_sample', 'reads', 'preproc']].sort_values('reads'), headers='keys', tablefmt='fancy_grid'))
# construct sample set name for fs
def_name = kwargs.pop('timepoint', None)
if def_name is None:
curr_date = datetime.datetime.now()
def_name = '{date}{month}{year}_{hour}{minute}%s'.format(
date=curr_date.strftime("%d"),
month=curr_date.strftime("%b"),
year=curr_date.strftime("%y"),
hour=curr_date.strftime("%H"),
minute=curr_date.strftime("%M"))
else:
def_name = '{def_name}%s'.format(def_name=def_name)
if meta_column is None and column_value is None:
sample_set_name = def_name % ('')
elif meta_column is not None and column_value is None:
sample_set_name = def_name % ('__' + meta_column)
else:
sample_set_name = def_name % ('__' + meta_column + '_' + column_value)
return sample_set, sample_set_name
def generic_command_individual_samples(config, df, preproc, meta_column,
column_value, samples_to_add,
exclude_samples, **kwargs):
"""
Construct sample sets, has multiple options.
Returns dict or sample_sets based on the provided options.
meta_column - factor column in sample metadata sheet. Cannot be combined with samples_to_add. exclude_samples has higher proirity.
column_value - value of column to select by. \
Can be multiple - separated by commas without whitespace. \
If meta_column is provided, bot no column_value is provided \
- treat it like select all unique values of that column.
If multiple - one value - one sample_set. If --merge enabled - all values go in one sample_set
assnake result request megahit -d FMT_FHM -c source run
"""
exclude_samples = [] if exclude_samples == '' else [
c.strip() for c in exclude_samples.split(',')
]
samples_to_add = [] if samples_to_add == '' else [
c.strip() for c in samples_to_add.split(',')
]
df_loaded = assnake.Dataset(df)
config['requested_dfs'] += [df_loaded.df]
# Now for the meta column stuff
meta_loc = os.path.join(df_loaded.full_path, 'meta.tsv')
if os.path.isfile(meta_loc):
meta = pd.read_csv(meta_loc, sep='\t')
if meta_column is not None:
if column_value is not None:
subset_by_col_value = meta.loc[meta[meta_column] ==
column_value]
if len(subset_by_col_value) > 0:
samples_to_add = list(subset_by_col_value['df_sample'])
if preproc is None:
# LONGEST
click.echo('Preprocessing is not specified, using longest for now')
preproc = max(list(df_loaded.sample_sets.keys()), key=len)
sample_set = assnake.api.loaders.load_sample_set(
config['wc_config'],
df_loaded.fs_prefix,
df_loaded.df,
preproc,
samples_to_add=samples_to_add)
if len(exclude_samples) > 0:
sample_set = sample_set.loc[
~sample_set['df_sample'].isin(exclude_samples), ]
# click.echo(tabulate(sample_set[['df_sample', 'reads', 'preproc']].sort_values('reads'), headers='keys', tablefmt='fancy_grid'))
# construct sample set name for fs
if meta_column is None and column_value is None:
# if only one sample is present in sample set, construct sample set name as `df_sample__{df_sample}`
print(sample_set)
if len(sample_set['df_sample']) == 1:
sample_set_name = 'df_sample__' + list(sample_set['df_sample'])[0]
else:
curr_date = datetime.datetime.now()
def_name = '{day}_{month}_{year}'.format(
month=curr_date.strftime("%b"),
year=curr_date.year,
day=curr_date.day)
sample_set_name = def_name
else:
sample_set_name = meta_column + '__' + column_value
return sample_set, sample_set_name