added generate inputs test, cleaned up code

src/analysis/summary.py

@@ -68,7 +68,7 @@ def summarize_networks(file_paths: Iterable[Path], node_table: pd.DataFrame) ->
 def degree(g):
     return dict(g.degree)
 
-
+# TODO: redo .run code to work on mixed graphs
 # stats is just a list of functions to apply to the graph.
 # They should take as input a networkx graph or digraph but may have any output.
 # stats = [degree, nx.clustering, nx.betweenness_centrality]

src/dataset.py

@@ -71,20 +71,10 @@ def load_files_from_dict(self, dataset_dict):
         data_loc = dataset_dict["data_dir"]
 
         # Load everything as pandas tables
-        print("about to create self.interactome")
-        print(data_loc)
-        print(interactome_loc)
-
-        with open(os.path.join(data_loc, interactome_loc), "r") as f:
-            for _i in range(9):  # first 5 lines
-                print(f.readline())
-
         self.interactome = pd.read_table(
             os.path.join(data_loc, interactome_loc), sep="\t", header=None
         )
-        print(self.interactome)
         num_cols = self.interactome.shape[1]
-        print(num_cols)
         if num_cols == 3:
 
             self.interactome.columns = ["Interactor1", "Interactor2", "Weight"]

src/domino.py

@@ -206,7 +206,6 @@ def parse_output(raw_pathway_file, standardized_pathway_file):
             edges_df['target'] = edges_df['target'].apply(post_domino_id_transform)
             edges_df = readd_direction_col_undirected(edges_df)
 
-        print(edges_df)
         edges_df.to_csv(standardized_pathway_file, sep='\t', header=False, index=False)
 
 

test/PrepareInputs/expected/ap-network-expected.txt

@@ -0,0 +1,3 @@
+#Interactor1	Interactor2	Weight
+A	B	0.98
+B	C	0.77

test/PrepareInputs/expected/d-network-expected.txt

@@ -0,0 +1,3 @@
+ID_interactor_A	ppi	ID_interactor_B
+ENSG0A	ppi	ENSG0B
+ENSG0B	ppi	ENSG0C

test/PrepareInputs/expected/mcf-edges-expected.txt

@@ -0,0 +1,4 @@
+A	B	0.98
+B	C	0.77
+B	A	0.98
+C	B	0.77

test/PrepareInputs/expected/meo-edges-expected.txt

@@ -0,0 +1,2 @@
+A	(pp)	B	0.98
+B	(pp)	C	0.77

test/PrepareInputs/expected/oi1-network-expected.txt

@@ -0,0 +1,3 @@
+protein1	protein2	weight	directionality
+A	B	0.98	U
+B	C	0.77	U

test/PrepareInputs/expected/oi2-network-expected.txt

@@ -0,0 +1,3 @@
+protein1	protein2	cost
+A	B	0.52
+B	C	0.73

test/PrepareInputs/expected/pl-network-expected.txt

@@ -0,0 +1,5 @@
+#Interactor1	Interactor2	Weight
+A	B	0.98
+B	C	0.77
+B	A	0.98
+C	B	0.77

test/PrepareInputs/test_prepare_inputs.py

@@ -0,0 +1,73 @@
+import filecmp
+import os
+import shutil
+from pathlib import Path
+
+import yaml
+
+from src import runner
+
+OUTDIR = "test/PrepareInputs/output/"
+EXPDIR = "test/PrepareInputs/expected/"
+class TestPrepareInputs:
+    def setup_class(cls):
+        """
+        Create the expected output directory
+        """
+        Path(OUTDIR).mkdir(parents=True, exist_ok=True)
+
+    def test_prepare_inputs(self):
+        config_loc = os.path.join("config", "config.yaml")
+
+        with open(config_loc) as config_file:
+            config = yaml.load(config_file, Loader=yaml.FullLoader)
+        test_file = "test/PrepareInputs/output/test_pickled_dataset.pkl"
+
+        for dataset in config["datasets"]:
+            runner.merge_input(dataset, test_file)
+
+            # Test PathLinker
+            filename_map = {"nodetypes": os.path.join("test","PrepareInputs", "output", "pl-nodetypes.txt"),
+                            "network": os.path.join("test","PrepareInputs", "output", "pl-network.txt")}
+            runner.prepare_inputs("pathlinker", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'pl-network.txt', EXPDIR + 'pl-network-expected.txt')
+
+            # Test OmicsIntegrator1
+            filename_map = {"prizes": os.path.join("test","PrepareInputs", "output", "oi1-prizes.txt"),
+                            "edges": os.path.join("test","PrepareInputs", "output", "oi1-network.txt")}
+            runner.prepare_inputs("omicsintegrator1", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'oi1-network.txt', EXPDIR + 'oi1-network-expected.txt')
+
+            # Test OmicsIntegrator2
+            filename_map = {"prizes": os.path.join("test","PrepareInputs", "output", "oi2-prizes.txt"),
+                            "edges": os.path.join("test","PrepareInputs", "output", "oi2-network.txt")}
+            runner.prepare_inputs("omicsintegrator2", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'oi2-network.txt', EXPDIR + 'oi2-network-expected.txt')
+
+            # Test MEO
+            filename_map = {"sources": os.path.join("test","PrepareInputs", "output", "meo-sources.txt"),
+                            "targets": os.path.join("test","PrepareInputs", "output", "meo-targets.txt"),
+                            "edges": os.path.join("test","PrepareInputs", "output", "meo-edges.txt")}
+            runner.prepare_inputs("meo", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'meo-edges.txt', EXPDIR + 'meo-edges-expected.txt')
+
+            # Test MCF
+            filename_map = {"sources": os.path.join("test","PrepareInputs", "output", "mcf-sources.txt"),
+                            "targets": os.path.join("test","PrepareInputs", "output", "mcf-targets.txt"),
+                            "edges": os.path.join("test","PrepareInputs", "output", "mcf-edges.txt")}
+            runner.prepare_inputs("mincostflow", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'mcf-edges.txt', EXPDIR + 'mcf-edges-expected.txt')
+
+            # Test AllPairs
+            filename_map = {"nodetypes": os.path.join("test","PrepareInputs", "output", "ap-nodetypes.txt"),
+                            "network": os.path.join("test","PrepareInputs", "output", "ap-network.txt")}
+            runner.prepare_inputs("allpairs", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'ap-network.txt', EXPDIR + 'ap-network-expected.txt')
+
+            # Test Domino
+            filename_map = {"network": os.path.join("test","PrepareInputs", "output", "d-network.txt"),
+                            "active_genes": os.path.join("test","PrepareInputs", "output", "d-active_genes.txt")}
+            runner.prepare_inputs("domino", test_file, filename_map)
+            assert filecmp.cmp(OUTDIR +'d-network.txt', EXPDIR + 'd-network-expected.txt')
+
+            break # only run dataset 0