Switch to tab delimited output files, with semi-colons for further di…

…visions

.gitignore

@@ -22,5 +22,5 @@
 # The results of the pytest cases #
 search.hits
 assignments.txt
-otus_by_rank.csv
-otus_cumulative.csv
+otus_by_rank.tsv
+otus_cumulative.tsv

crest4/classify.py

@@ -330,10 +330,12 @@ def __call__(self):
         self.out_file.writelines(query.tax_string for query in self.queries)
         # Special case where an OTU table was passed #
         if self.otu_table:
-            otus_by_rank    = self.output_dir + 'otus_by_rank.csv'
-            otus_cumulative = self.output_dir + 'otus_cumulative.csv'
-            self.otu_info(cumulative=False).to_csv(otus_by_rank, index=False)
-            self.otu_info(cumulative=True).to_csv(otus_cumulative, index=False)
+            path_by_rank    = self.output_dir + 'otus_by_rank.tsv'
+            path_cumulative = self.output_dir + 'otus_cumulative.tsv'
+            otus_by_rank    = self.otu_info.otus_by_rank
+            otus_cumulative = self.otu_info.otus_cumulative
+            otus_by_rank.to_csv(path_by_rank, index=False, sep='\t')
+            otus_cumulative.to_csv(path_cumulative, index=False, sep='\t')
         # Print a success message #
         msg = "Classification ran successfully. Results are placed in '%s'."
         print(msg % self.out_file)

crest4/otu_tables.py

@@ -95,22 +95,22 @@ def __call__(self, cumulative=False):
             # By default the root is at the end
             tax = list(reversed(tax))
             # Make a string #
-            tax_name = ' / '.join(tax)
+            tax_name = '; '.join(tax)
             # Add the current counts to that particular taxonomy #
             result[tax_name] += otu_counts
             # If we have the cumulative option then propagate up the tree #
             if cumulative:
                 for step in range(1, len(tax)):
-                    tax_name = ' / '.join(tax[0:-step])
+                    tax_name = '; '.join(tax[0:-step])
                     result[tax_name] += otu_counts
         # Convert to a DataFrame #
         result = pandas.DataFrame(result).T
         # Have the assignment as a separate column #
         result = result.reset_index()
         result = result.rename(columns = {'index': 'taxonomy'})
-        # Add the rank column that tells you if it's a genus or a family #
+        # Add the rank column that tells the user if it's a genus or a family #
         rank_names  = self.classify.database.rank_names
-        tax_to_rank = lambda t: rank_names[len(t.split('/')) - 1]
+        tax_to_rank = lambda t: rank_names[len(t.split(';')) - 1]
         ranks       = result.taxonomy.apply(tax_to_rank)
         result.insert(loc=0, column='rank', value=ranks)
         # Sort the table by the taxonomy string #

crest4/query.py

@@ -147,7 +147,7 @@ def tax_string(self):
         the full assigned taxonomy along with the original name of the
         sequence classified.
         """
-        # Make a comma separated string #
-        tax = ','.join(reversed(self.taxonomy))
+        # Make a semi colon separated string #
+        tax = '; '.join(reversed(self.taxonomy))
         # Add the name of the query to the beginning line #
         return self.name + '\t' + tax + '\n'