def test_keep_columns_from_mapping_file(self):
"""Check correct selection of metadata is being done"""
# test it returns the same data
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers, [])
self.assertEquals(out_data, [[], [], [], [], [], [], [], [], []])
self.assertEquals(out_headers, [])
# test it can filter a list of columns
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers,
['SampleID', 'LinkerPrimerSequence', 'Description'])
self.assertEquals(out_headers,
['SampleID', 'LinkerPrimerSequence', 'Description'])
self.assertEquals(out_data, PRE_PROCESS_B)
# test correct negation of filtering
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers,
['LinkerPrimerSequence', 'Description'], True)
self.assertEquals(out_data, PRE_PROCESS_A)
self.assertEquals(out_headers,
['SampleID', 'BarcodeSequence', 'Treatment', 'DOB'])
def test_keep_columns_from_mapping_file(self):
"""Check correct selection of metadata is being done"""
# test it returns the same data
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers, [])
self.assertEquals(out_data, [[], [], [], [], [], [], [], [], []])
self.assertEquals(out_headers, [])
# test it can filter a list of columns
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers,
['SampleID', 'LinkerPrimerSequence', 'Description'])
self.assertEquals(out_headers,
['SampleID', 'LinkerPrimerSequence', 'Description'])
self.assertEquals(out_data, PRE_PROCESS_B)
# test correct negation of filtering
out_data, out_headers = keep_columns_from_mapping_file(
self.mapping_file_data, self.mapping_file_headers,
['LinkerPrimerSequence', 'Description'], True)
self.assertEquals(out_data, PRE_PROCESS_A)
self.assertEquals(out_headers,
['SampleID', 'BarcodeSequence', 'Treatment', 'DOB'])
def format_vectors_to_js(mapping_file_data, mapping_file_headers, coords_data,
coords_headers, connected_by_header,
sorted_by_header=None):
"""Write a string representing the vectors in a PCoA plot as javascript
Inputs:
mapping_file_data: contents of the mapping file
mapping_file_headers: headers of the mapping file
coords_data: coordinates of the PCoA plot in a numpy 2-D array or a list of
numpy 2-D arrays for jackknifed input
coords_headers: headers of the coords in the PCoA plot or a list of lists
with the headers for jackknifed input
connected_by_header: header of the mapping file that represents how the
lines will be connected
sorted_by_header: numeric-only header name to sort the samples in the
vectors
Output:
js_vectors_string: string that represents the vectors in the shape of a
javascript object
Notes:
If using jackknifed input, the coordinates and headers that will be used are
the ones belonging to the master coords i. e. the first element.
"""
js_vectors_string = []
js_vectors_string.append('\nvar g_vectorPositions = new Array();\n')
if connected_by_header != None:
# check if we are processing jackknifed input, if so just get the master
if type(coords_data) == list:
coords_data = coords_data[0]
coords_headers = coords_headers[0]
columns_to_keep = ['SampleID', connected_by_header]
# do not ad None if sorted_by_header is None or empty
if sorted_by_header:
columns_to_keep.append(sorted_by_header)
# reduce the amount of data by keeping the required fields only
mapping_file_data, mapping_file_headers =\
keep_columns_from_mapping_file(mapping_file_data,
mapping_file_headers, columns_to_keep)
# format the mapping file to use this with the filtering function
mf_string = format_mapping_file(mapping_file_headers, mapping_file_data)
index = mapping_file_headers.index(connected_by_header)
connected_by = list(set([line[index] for line in mapping_file_data]))
for category in connected_by:
# convert to StringIO to for each iteration; else the object
# won't be usable after the first iteration & you'll get an error
sample_ids = sample_ids_from_metadata_description(
StringIO(mf_string),'%s:%s' % (connected_by_header,category))
# if there is a sorting header, sort the coords using these values
if sorted_by_header:
sorting_index = mapping_file_headers.index(sorted_by_header)
to_sort = [line for line in mapping_file_data if line[0] in\
sample_ids]
# get the sorted sample ids from the sorted-reduced mapping file
sample_ids = zip(*sorted(to_sort,
key=lambda x: float(x[sorting_index])))[0]
# each category value is a new vector
js_vectors_string.append("g_vectorPositions['%s'] = new Array();\n"
% (category))
for s in sample_ids:
index = coords_headers.index(s)
# print the first three elements of each coord for each sample
js_vectors_string.append("g_vectorPositions['%s']['%s'] = %s;\n"
% (category, s, coords_data[index, :3].tolist()))
return ''.join(js_vectors_string)
def format_vectors_to_js(mapping_file_data,
mapping_file_headers,
coords_data,
coords_headers,
connected_by_header,
sorted_by_header=None):
"""Write a string representing the vectors in a PCoA plot as javascript
Inputs:
mapping_file_data: contents of the mapping file
mapping_file_headers: headers of the mapping file
coords_data: coordinates of the PCoA plot in a numpy 2-D array or a list of
numpy 2-D arrays for jackknifed input
coords_headers: headers of the coords in the PCoA plot or a list of lists
with the headers for jackknifed input
connected_by_header: header of the mapping file that represents how the
lines will be connected
sorted_by_header: numeric-only header name to sort the samples in the
vectors
Output:
js_vectors_string: string that represents the vectors in the shape of a
javascript object
Notes:
If using jackknifed input, the coordinates and headers that will be used are
the ones belonging to the master coords i. e. the first element.
"""
js_vectors_string = []
js_vectors_string.append('\nvar g_vectorPositions = new Array();\n')
if connected_by_header != None:
# check if we are processing jackknifed input, if so just get the master
if type(coords_data) == list:
coords_data = coords_data[0]
coords_headers = coords_headers[0]
columns_to_keep = ['SampleID', connected_by_header]
# do not ad None if sorted_by_header is None or empty
if sorted_by_header:
columns_to_keep.append(sorted_by_header)
# reduce the amount of data by keeping the required fields only
mapping_file_data, mapping_file_headers =\
keep_columns_from_mapping_file(mapping_file_data,
mapping_file_headers, columns_to_keep)
# format the mapping file to use this with the filtering function
mf_string = format_mapping_file(mapping_file_headers,
mapping_file_data)
index = mapping_file_headers.index(connected_by_header)
connected_by = list(set([line[index] for line in mapping_file_data]))
for category in connected_by:
# convert to StringIO to for each iteration; else the object
# won't be usable after the first iteration & you'll get an error
sample_ids = sample_ids_from_metadata_description(
StringIO(mf_string), '%s:%s' % (connected_by_header, category))
# if there is a sorting header, sort the coords using these values
if sorted_by_header:
sorting_index = mapping_file_headers.index(sorted_by_header)
to_sort = [line for line in mapping_file_data if line[0] in\
sample_ids]
# get the sorted sample ids from the sorted-reduced mapping file
sample_ids = zip(
*sorted(to_sort, key=lambda x: float(x[sorting_index])))[0]
# each category value is a new vector
js_vectors_string.append(
"g_vectorPositions['%s'] = new Array();\n" % (category))
for s in sample_ids:
index = coords_headers.index(s)
# print the first three elements of each coord for each sample
js_vectors_string.append(
"g_vectorPositions['%s']['%s'] = %s;\n" %
(category, s, coords_data[index, :3].tolist()))
return ''.join(js_vectors_string)
