Implemented PhenotypeDataset with tests.

system/dataset.py

@@ -56,7 +56,7 @@ class Dataset:
             if shape != None:
                 if self.shape!=shape:
                     #logger.log("debug","Dataset and input shape mismatch")
-                    raise ValueError
+                    raise ValueError, "Differing in array and provided. %s != %s" % (self.shape, shape)
             if identifiers!=None:
                 self._set_identifiers(identifiers,all_dims)
             else:

test/workflows/affy_workflowtest.py

@@ -0,0 +1,81 @@
+import unittest
+import sys
+
+sys.path.append("../..")
+from workflows.affy_workflow import PhenotypeDataset
+
+class PhenotypeDatasetTest(unittest.TestCase):
+    def testEmptyData(self):
+        # we have a list of cel-files, but no categories
+        cel_data = """\
+CEL
+02-05-33
+03-07-38
+"""
+        dataset = PhenotypeDataset(cel_data)
+        self.assertEquals(['CEL', 'phenotypes'], dataset.get_dim_names())
+        self.assertEquals(['02-05-33', '03-07-38'], dataset.get_identifiers('CEL'))
+        self.assertEquals([], dataset.get_identifiers('phenotypes'))
+
+    def testFloatData(self):
+        cel_data = """\
+CEL\tage
+02-05-33\t8
+03-07-38\t9
+"""
+        dataset = PhenotypeDataset(cel_data)
+        self.assertEquals(['CEL', 'phenotypes'], dataset.get_dim_names())
+        self.assertEquals(['age'], dataset.get_identifiers('phenotypes'))
+        self.assertEquals([[8],
+                           [9]], dataset.asarray().tolist())
+
+    def testCategoryData(self):
+        """Categories expand to one numeric column for each category choice."""
+        cel_data = """\
+CEL\tsick
+02-05-33\tyes
+03-07-38\tno
+04-93-33\tyes
+08-32-33\tmaybe
+"""
+        dataset = PhenotypeDataset(cel_data)
+        self.assertEquals(['CEL', 'phenotypes'], dataset.get_dim_names())
+        self.assertEquals(['sick-yes', 'sick-no', 'sick-maybe'], dataset.get_identifiers('phenotypes'))
+        self.assertEquals([[1, 0, 0],
+                           [0, 1, 0],
+                           [1, 0, 0],
+                           [0, 0, 1]], dataset.asarray().tolist())
+
+    def testMultipleCategoriesAndFloats(self):
+        cel_data = """\
+CEL\tsex\tage\tinfected
+02-05-33\tF\t8\tI
+02-05-34\tF\t9\tN
+02-05-35\tM\t8\tI
+"""
+        dataset = PhenotypeDataset(cel_data)
+        self.assertEquals(['sex-F', 'sex-M', 'age', 'infected-I', 'infected-N'],
+                          dataset.get_identifiers('phenotypes'))
+
+        self.assertEquals([[1, 0, 8, 1, 0],
+                           [1, 0, 9, 0, 1],
+                           [0, 1, 8, 1, 0]], dataset.asarray().tolist())
+
+    def testGetPhenotypeTable(self):
+        cel_data = """\
+CEL\tsex\tage\tinfected
+02-05-33\tF\t8\tI
+02-05-34\tF\t9\tN
+02-05-35\tM\t8\tI
+"""
+        dataset = PhenotypeDataset(cel_data)
+
+        self.assertEquals([['CEL', 'sex', 'age', 'infected'],
+                           ['02-05-33', 'F', '8', 'I'],
+                           ['02-05-34', 'F', '9', 'N'],
+                           ['02-05-35', 'M', '8', 'I']], dataset.get_phenotype_table())
+
+
+
+if __name__=='__main__':
+    unittest.main()

workflows/affy_workflow.py

@@ -1,6 +1,6 @@
 import gtk
 from system import dataset, logger, plots, workflow
-from scipy import randn
+from scipy import randn, array, transpose, zeros
 import cPickle
 
 
@@ -185,3 +185,53 @@ class PCAFunction(workflow.Function):
         
         return [T, P, loading_plot, score_plot]
 
+
+class PhenotypeDataset(dataset.Dataset):
+    def __init__(self, string):
+        self._table = rows = [line.split("\t") for line in string.splitlines()]
+        columns = zip(*rows[1:])
+        cel_names = columns[0]
+        col_names = rows[0][1:]
+        phenotypes = []
+        categories = {}
+
+        for col_name, column in zip(col_names, columns[1:]):
+            try:
+                categories[col_name] = map(int, column)
+                phenotypes.append(col_name)
+            except ValueError:
+                # category-data
+                keys = []
+                entries = {}
+                for i, entry in enumerate(column):
+                    if entry not in entries:
+                        keys.append(entry)
+                        entries[entry] = []
+
+                    entries[entry].append(i)
+                    
+
+                for key in keys:
+                    z = zeros(len(column))
+                    for i in entries[key]:
+                        z[i] = 1
+                    key = "%s-%s" % (col_name, key)
+                    phenotypes.append(key)
+                    categories[key] = z
+
+        matrix_data = []
+        for key in phenotypes:
+            matrix_data.append(categories[key])
+
+        if matrix_data:
+            a = transpose(array(matrix_data))
+        else:
+            a = None
+
+        dataset.Dataset.__init__(self, a, identifiers=[('CEL', cel_names),
+                                                       ('phenotypes', phenotypes)],
+                                 shape=(len(cel_names),len(phenotypes)))
+            
+
+    def get_phenotype_table(self):
+        return self._table