def test_sort_taxa_table_by_pcoa_coords(self):
"""Make sure OTU table and coordinates are sorted equally"""
# case with shuffled inputs
o_headers, o_otu_table = sort_taxa_table_by_pcoa_coords(
self.coords_header, self.otu_table, self.otu_headers)
self.assertEquals(o_headers, [
'PC.354', 'PC.356', 'PC.481', 'PC.593', 'PC.355', 'PC.607',
'PC.634', 'PC.636', 'PC.635'
])
assert_almost_equal(o_otu_table, OTU_TABLE_A)
# case with shuffled inputs and fewer samples
o_headers, o_otu_table = sort_taxa_table_by_pcoa_coords(
['PC.354', 'PC.356', 'PC.635'], self.otu_table, self.otu_headers)
self.assertEquals(o_headers, ['PC.354', 'PC.356', 'PC.635'])
assert_almost_equal(
o_otu_table,
array([[0.01, 0.02, 0.04697987], [0., 0.02, 0.02013423],
[0.38926174, 0.65333333, 0.27516779],
[0., 0.03333333, 0.02013423],
[0.41610738, 0.22, 0.45637584],
[0.03355705, 0.01333333, 0.02013423], [0., 0.01333333, 0.],
[0.14765101, 0.02666667, 0.16107383]]))
def test_sort_taxa_table_by_pcoa_coords(self):
"""Make sure OTU table and coordinates are sorted equally"""
# case with shuffled inputs
o_headers, o_otu_table = sort_taxa_table_by_pcoa_coords(
self.coords_header, self.otu_table, self.otu_headers)
self.assertEquals(o_headers, ['PC.354', 'PC.356', 'PC.481', 'PC.593',
'PC.355', 'PC.607', 'PC.634', 'PC.636',
'PC.635'])
assert_almost_equal(o_otu_table, OTU_TABLE_A)
# case with shuffled inputs and fewer samples
o_headers, o_otu_table = sort_taxa_table_by_pcoa_coords(
['PC.354', 'PC.356', 'PC.635'], self.otu_table, self.otu_headers)
self.assertEquals(o_headers, ['PC.354', 'PC.356', 'PC.635'])
assert_almost_equal(o_otu_table, array(
[[0.01, 0.02, 0.04697987],
[0., 0.02, 0.02013423],
[0.38926174, 0.65333333, 0.27516779],
[0., 0.03333333, 0.02013423],
[0.41610738, 0.22, 0.45637584],
[0.03355705, 0.01333333, 0.02013423],
[0., 0.01333333, 0.],
[0.14765101, 0.02666667, 0.16107383]]))
def preprocess_otu_table(otu_sample_ids,
otu_table,
lineages,
coords_data,
coords_headers,
N=0):
"""Preprocess the OTU table to to generate the required data for the biplots
Input:
otu_sample_ids: sample identifiers for the otu_table
otu_table: contingency table
lineages: taxonomic assignments for the OTUs in the otu_table
coords_data: principal coordinates data where the taxa will be mapped
N: number of most prevalent taxa to keep, by default will use all
Output:
otu_coords: coordinates representing the N most prevalent taxa in otu_table
otu_table: N most prevalent OTUs from the input otu_table
otu_lineages: taxonomic assignments corresponding to the N most prevalent
OTUs
otu_prevalence: vector with the prevalence scores of the N highest values
lines: coords where the N most prevalent taxa will be positioned in the
biplot
"""
# return empty values if any of the taxa data is empty
if (otu_sample_ids == []) or (otu_table == array([])) or (lineages == []):
return [], [], [], [], ''
# this means there's only one or fewer rows in the contingency table
if len(otu_table) <= 1 or len(lineages) <= 1:
raise EmperorUnsupportedComputation, "Biplots are not supported for "+\
"contingency tables with one or fewer rows"
# if this element is a list take the first headers and coordinates
# both of these will be the master coordinates, i. e. where data is centered
if type(coords_data) == list and type(coords_headers) == list:
coords_data = coords_data[0]
coords_headers = coords_headers[0]
# re-arrange the otu table so it matches the order of the samples in the
# coordinates data & remove any sample that is not in the coordinates header
otu_sample_ids, otu_table = sort_taxa_table_by_pcoa_coords(
coords_headers, otu_table, otu_sample_ids)
# retrieve the prevalence and the coords prior the filtering
prevalence = get_taxa_prevalence(otu_table)
bi_plot_coords = get_taxa_coords(otu_table, coords_data)
o_otu_coords, o_otu_table, o_otu_lineages, o_prevalence =\
extract_taxa_data(bi_plot_coords, otu_table, lineages, prevalence, N)
lines = '\n'.join(
make_biplot_scores_output({
'coord': o_otu_coords,
'lineages': o_otu_lineages
}))
return o_otu_coords, o_otu_table, o_otu_lineages, o_prevalence, lines
def preprocess_otu_table(otu_sample_ids, otu_table, lineages,
coords_data, coords_headers, N=0):
"""Preprocess the OTU table to to generate the required data for the biplots
Input:
otu_sample_ids: sample identifiers for the otu_table
otu_table: contingency table
lineages: taxonomic assignments for the OTUs in the otu_table
coords_data: principal coordinates data where the taxa will be mapped
N: number of most prevalent taxa to keep, by default will use all
Output:
otu_coords: coordinates representing the N most prevalent taxa in otu_table
otu_table: N most prevalent OTUs from the input otu_table
otu_lineages: taxonomic assignments corresponding to the N most prevalent
OTUs
otu_prevalence: vector with the prevalence scores of the N highest values
lines: coords where the N most prevalent taxa will be positioned in the
biplot
"""
# return empty values if any of the taxa data is empty
if (otu_sample_ids == []) or (otu_table == array([])) or (lineages == []):
return [], [], [], [], ''
# this means there's only one or fewer rows in the contingency table
if len(otu_table) <= 1 or len(lineages) <= 1:
raise EmperorUnsupportedComputation, "Biplots are not supported for "+\
"contingency tables with one or fewer rows"
# if this element is a list take the first headers and coordinates
# both of these will be the master coordinates, i. e. where data is centered
if type(coords_data) == list and type(coords_headers) == list:
coords_data = coords_data[0]
coords_headers = coords_headers[0]
# re-arrange the otu table so it matches the order of the samples in the
# coordinates data & remove any sample that is not in the coordinates header
otu_sample_ids, otu_table = sort_taxa_table_by_pcoa_coords(coords_headers,
otu_table, otu_sample_ids)
# retrieve the prevalence and the coords prior the filtering
prevalence = get_taxa_prevalence(otu_table)
bi_plot_coords = get_taxa_coords(otu_table, coords_data)
o_otu_coords, o_otu_table, o_otu_lineages, o_prevalence =\
extract_taxa_data(bi_plot_coords, otu_table, lineages, prevalence, N)
lines = '\n'.join(make_biplot_scores_output({'coord': o_otu_coords,
'lineages': o_otu_lineages}))
return o_otu_coords, o_otu_table, o_otu_lineages, o_prevalence, lines
