def writesorted(self, lineages, base_file, path, metric='euclidean'):
'''
Do hierarchical clustering of rows and cols and write the sorted data to files.
'''
from heatmap_clust import clust_data, heatmap_clust
matrix, row_labels, col_labels = self.to_matrix()
mat_sorted, row_labels_sorted, col_labels_sorted = clust_data(
matrix,
metric,
row_labels=row_labels,
col_labels=col_labels,
row_label_width=.25)
row_labels_sorted.reverse()
mat_file = path + base_file + '_abunds_sorted.txt'
np.savetxt(mat_file, mat_sorted)
otu_file = path + base_file + '_OTUs_sorted.txt'
f = open(otu_file, 'w')
for otu in row_labels_sorted:
f.write('\t'.join([otu, lineages[otu]]) + '\n')
f.close()
samples_file = path + base_file + '_samples_sorted.txt'
f = open(samples_file, 'w')
for sample in col_labels_sorted:
f.write(sample + '\n')
f.close()
def plot_heatmap(self, row_metric ='euclidean', col_metric ='euclidean', file = None, **kwargs):
'''
Plot heatmap of self, sorted by heirarchical clustering with given distance metric
'''
from heatmap_clust import clust_data, heatmap_clust
matrix, row_labels, col_labels = self.to_matrix()
plot_row_labels = kwargs.get('plot_row_labels',False)
plot_col_labels = kwargs.get('plot_col_labels',False)
if plot_row_labels and plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, row_labels = row_labels, col_labels = col_labels, **kwargs)
elif plot_row_labels and not plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, row_labels = row_labels, **kwargs)
elif not plot_row_labels and plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, col_labels = col_labels, **kwargs)
else:
clust_data(matrix, row_metric, col_metric, file = file, **kwargs)
def plot_heatmap(self, row_metric ='euclidean', col_metric ='euclidean', file = None, **kwargs):
'''
Plot heatmap of self, sorted by heirarchical clustering with given distance metric
'''
from heatmap_clust import clust_data, heatmap_clust
matrix, row_labels, col_labels = self.to_matrix()
plot_row_labels = kwargs.get('plot_row_labels',False)
plot_col_labels = kwargs.get('plot_col_labels',False)
if plot_row_labels and plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, row_labels = row_labels, col_labels = col_labels, **kwargs)
elif plot_row_labels and not plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, row_labels = row_labels, **kwargs)
elif not plot_row_labels and plot_col_labels:
clust_data(matrix, row_metric, col_metric, file = file, col_labels = col_labels, **kwargs)
else:
clust_data(matrix, row_metric, col_metric, file = file, **kwargs)
def writesorted(self, lineages, base_file, path, metric = 'euclidean'):
'''
Do hierarchical clustering of rows and cols and write the sorted data to files.
'''
from heatmap_clust import clust_data, heatmap_clust
matrix, row_labels, col_labels = self.to_matrix()
mat_sorted, row_labels_sorted, col_labels_sorted = clust_data(matrix, metric, row_labels = row_labels,col_labels = col_labels, row_label_width = .25)
row_labels_sorted.reverse()
mat_file = path + base_file + '_abunds_sorted.txt'
np.savetxt(mat_file, mat_sorted)
otu_file = path + base_file + '_OTUs_sorted.txt'
f = open(otu_file,'w')
for otu in row_labels_sorted:
f.write('\t'.join([otu, lineages[otu]]) + '\n')
f.close()
samples_file = path + base_file + '_samples_sorted.txt'
f = open(samples_file,'w')
for sample in col_labels_sorted:
f.write(sample + '\n')
f.close()
