def generate_alpha_rarefaction_data_from_point_in_omega(
biom_object, metrics, sequences, iterations, tree_object=None):
"""generate alpha rarefaction data from a biom table and mapping file
Inputs:
biom_object: OTU table to be rarefied and used to compute alpha diversity
metrics: list of metrics, phylogenetic or non phylogenetic
sequences: maximum number of sequences for the rarefaction plots
iterations: number of repetitions per rarefaction
tree_object: tree to perform the phylogenetic operations, default is None
Output:
alpha_rarefaction_data: dictionary where the keys are alpha diversity
metrics and the values are tuples; in these tuples the first element is a
list of column headers for an alpha diversity file, the second element is a
list of row headers for an alpha diversity file and the third element is a
list of lists containing the alpha diversity data computed at multiple
rarefaction depths and as many iterations as specified.
"""
# The minimum depth is defined by the size of the maximum depth
steps = 4
min_depth = int(ceil(sequences / steps))
# get a rarefied biom with the proper identifiers
rarefied_bioms_list = get_rarefactions(biom_object, min_depth, sequences,\
iterations, steps)
alpha_rs = {}
alpha_filenames = []
# rarefy all the biom objects and get the alpha diversity values
for rarefied_biom in rarefied_bioms_list:
# this tag contains data about the iteration and the depth
identifier = 'alpha_rare_%s_%s' % (str(
rarefied_biom[0]), str(rarefied_biom[1]))
alpha_values = single_object_alpha(rarefied_biom[2], metrics,
tree_object)
alpha_rs[identifier] = (rarefied_biom[0], rarefied_biom[1],
alpha_values.split('\n'))
alpha_filenames.append(identifier)
# use the rarefaction with the fewest sequences per sample as the reference
ref_rare = single_object_alpha(rarefied_bioms_list[0][2], metrics,\
tree_object=tree_object).split('\n')
all_metrics, all_samples, example_data = parse_matrix(ref_rare)
# build a dictionary with the data for each of the metrics specified
metrics_data = {}
for metric in all_metrics:
per_metric_data = []
for filename in alpha_filenames:
f_metrics, f_samples, f_data = parse_matrix(alpha_rs[filename][2])
per_metric_data.append(make_output_row(f_metrics, metric,\
f_samples, f_data, filename, len(all_samples), all_samples))
metrics_data[metric] = per_metric_data
# now format the dictionary to make it compatible with make_averages
alpha_rarefaction_data = _format_rarefactions(metrics_data, all_samples)
return alpha_rarefaction_data
def generate_alpha_rarefaction_data_from_point_in_omega(biom_object, metrics,
sequences, iterations,
tree_object=None):
"""generate alpha rarefaction data from a biom table and mapping file
Inputs:
biom_object: OTU table to be rarefied and used to compute alpha diversity
metrics: list of metrics, phylogenetic or non phylogenetic
sequences: maximum number of sequences for the rarefaction plots
iterations: number of repetitions per rarefaction
tree_object: tree to perform the phylogenetic operations, default is None
Output:
alpha_rarefaction_data: dictionary where the keys are alpha diversity
metrics and the values are tuples; in these tuples the first element is a
list of column headers for an alpha diversity file, the second element is a
list of row headers for an alpha diversity file and the third element is a
list of lists containing the alpha diversity data computed at multiple
rarefaction depths and as many iterations as specified.
"""
# The minimum depth is defined by the size of the maximum depth
steps = 4
min_depth = int(ceil(sequences / steps))
# get a rarefied biom with the proper identifiers
rarefied_bioms_list = get_rarefactions(biom_object, min_depth, sequences,\
iterations, steps)
alpha_rs = {}
alpha_filenames = []
# rarefy all the biom objects and get the alpha diversity values
for rarefied_biom in rarefied_bioms_list:
# this tag contains data about the iteration and the depth
identifier = 'alpha_rare_%s_%s' % (str(rarefied_biom[0]), str(rarefied_biom[1]))
alpha_values = single_object_alpha(rarefied_biom[2], metrics, tree_object)
alpha_rs[identifier] = (rarefied_biom[0], rarefied_biom[1], alpha_values.split('\n'))
alpha_filenames.append(identifier)
# use the rarefaction with the fewest sequences per sample as the reference
ref_rare = single_object_alpha(rarefied_bioms_list[0][2], metrics,\
tree_object=tree_object).split('\n')
all_metrics, all_samples, example_data = parse_matrix(ref_rare)
# build a dictionary with the data for each of the metrics specified
metrics_data = {}
for metric in all_metrics:
per_metric_data = []
for filename in alpha_filenames:
f_metrics, f_samples, f_data = parse_matrix(alpha_rs[filename][2])
per_metric_data.append(make_output_row(f_metrics, metric,\
f_samples, f_data, filename, len(all_samples), all_samples))
metrics_data[metric] = per_metric_data
# now format the dictionary to make it compatible with make_averages
alpha_rarefaction_data = _format_rarefactions(metrics_data, all_samples)
return alpha_rarefaction_data
def test_make_output_rows(self):
f_metrics = ["met1"]
metric = "met1"
f_samples = ["s1", "s2"]
f_data = numpy.array([[0.4], [0.8]])
fname = "alpha_rarefaction_10_7"
num_cols = 2
all_samples = ["s1", "s2"]
res = make_output_row(f_metrics, metric, f_samples, f_data, fname, num_cols, all_samples)
self.assertEqual(res, ["alpha_rarefaction_10_7", 10, 7, "0.4", "0.8"])
def test_make_output_rows(self):
f_metrics = ['met1']
metric = 'met1'
f_samples = ['s1','s2']
f_data = numpy.array([[.4],[.8]])
fname = 'alpha_rarefaction_10_7'
num_cols= 2
all_samples = ['s1','s2']
res = make_output_row(f_metrics, metric, f_samples,
f_data, fname, num_cols, all_samples)
self.assertEqual(res,['alpha_rarefaction_10_7', 10, 7, '0.4', '0.8'])
def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
if len(args) != 0:
parser.error("Positional argument detected. make sure all" +
' parameters are identified.' +
'\ne.g.: include the \"-m\" in \"-m MINIMUM_LENGTH\"')
input_dir = opts.input_path
output_dir = opts.output_path
example_filepath = opts.example_path
if not os.path.exists(output_dir):
os.makedirs(output_dir)
file_names = os.listdir(input_dir)
file_names = [fname for fname in file_names if not fname.startswith('.')]
if example_filepath is None:
# table row is base_name, seqs_per_sam, iters, ext
file_name_table = map(parse_rarefaction_fname, file_names)
# sort on seqs/sam
sorted_fname_table = sorted(
file_name_table,
key=operator.itemgetter(1))
# now map back to file name
example_fname = file_names[
file_name_table.index(sorted_fname_table[0])]
example_filepath = os.path.join(input_dir, example_fname)
f = open(example_filepath, 'U')
all_metrics, all_samples, example_data = parse_matrix(f)
num_cols = len(all_samples)
f.close()
# make the table 1 row at a time
# we're building a rarefaction by sample mtx from
# a sample by metric matrix
# each metric is one output file
for metric in all_metrics:
metric_file_data = []
for fname in file_names:
# f_ here refers to the input file currently being processed
# to distinguish from the output file we're building
f = open(os.path.join(input_dir, fname), 'U')
f_metrics, f_samples, f_data = parse_matrix(f)
f.close()
metric_file_data.append(
make_output_row(f_metrics, metric, f_samples,
f_data, fname, num_cols, all_samples))
write_output_file(metric_file_data, output_dir, metric, all_samples)
def main():
option_parser, opts, args = parse_command_line_parameters(**script_info)
if len(args) != 0:
parser.error("Positional argument detected. make sure all" +
' parameters are identified.' +
'\ne.g.: include the \"-m\" in \"-m MINIMUM_LENGTH\"')
input_dir = opts.input_path
output_dir = opts.output_path
example_filepath = opts.example_path
if not os.path.exists(output_dir):
os.makedirs(output_dir)
file_names = os.listdir(input_dir)
file_names = [fname for fname in file_names if not fname.startswith('.')]
if example_filepath is None:
# table row is base_name, seqs_per_sam, iters, ext
file_name_table = map(parse_rarefaction_fname, file_names)
# sort on seqs/sam
sorted_fname_table = sorted(
file_name_table,
key=operator.itemgetter(1))
# now map back to file name
example_fname = file_names[
file_name_table.index(sorted_fname_table[0])]
example_filepath = os.path.join(input_dir, example_fname)
f = open(example_filepath, 'U')
all_metrics, all_samples, example_data = parse_matrix(f)
num_cols = len(all_samples)
f.close()
# make the table 1 row at a time
# we're building a rarefaction by sample mtx from
# a sample by metric matrix
# each metric is one output file
for metric in all_metrics:
metric_file_data = []
for fname in file_names:
# f_ here refers to the input file currently being processed
# to distinguish from the output file we're building
f = open(os.path.join(input_dir, fname), 'U')
f_metrics, f_samples, f_data = parse_matrix(f)
f.close()
metric_file_data.append(
make_output_row(f_metrics, metric, f_samples,
f_data, fname, num_cols, all_samples))
write_output_file(metric_file_data, output_dir, metric, all_samples)
