fix: make dataset nodes ordering stable

docs/contributing/index.rst

@@ -338,11 +338,9 @@ image using the ``Dockerfile`` that was completed in Step 2.
 
 Modify generate inputs:
 
-1. Include a key-value pair in the algo_exp_file dictionary that links
-   the specific algorithm to its expected network file.
-2. Obtain the expected network file from the workflow, manually confirm
-   it is correct, and save it to ``test/generate-inputs/expected``. Name
-   it as ``{algorithm_name}-{network_file_name}-expected.txt``.
+1. Obtain the expected network and nodes file from the workflow,
+   manually confirm it is correct, and save it to ``test/generate-inputs/expected``.
+   Name it as ``{algorithm_name}-{network_file_name}-expected.txt``.
 
 Modify parse outputs:
 

spras/dataset.py

@@ -104,11 +104,12 @@ def load_files_from_dict(self, dataset_dict):
                 f"Edge file {interactome_loc} must have three or four columns but found {num_cols}"
             )
 
-        node_set = set(self.interactome.Interactor1.unique())
-        node_set = node_set.union(set(self.interactome.Interactor2.unique()))
+        # We get uniqueness afterwards to make `load_files_from_dict` have a well-defined node ordering,
+        # since algorithms may depend on the order of nodes passed.
+        nodes = list(pd.concat([self.interactome.Interactor1, self.interactome.Interactor2]).unique())
 
         # Load generic node tables
-        self.node_table = pd.DataFrame(node_set, columns=[self.NODE_ID])
+        self.node_table = pd.DataFrame(nodes, columns=[self.NODE_ID])
         for node_file in node_data_files:
             single_node_table = pd.read_table(os.path.join(data_loc, node_file))
             # If we have only 1 column, assume this is an indicator variable

test/generate-inputs/expected/allpairs/allpairs-directed_flag-expected.txt

@@ -0,0 +1 @@
+false

test/generate-inputs/expected/allpairs/allpairs-nodetypes-expected.txt

@@ -0,0 +1,3 @@
+#Node	Node type
+test_A	source
+C	target

test/generate-inputs/expected/bowtiebuilder/bowtiebuilder-sources-expected.txt

@@ -0,0 +1 @@
+test_A

test/generate-inputs/expected/bowtiebuilder/bowtiebuilder-targets-expected.txt

@@ -0,0 +1 @@
+C

test/generate-inputs/expected/domino/domino-active_genes-expected.txt

@@ -0,0 +1,2 @@
+ENSG0test_A
+ENSG0C

test/generate-inputs/expected/meo/meo-sources-expected.txt

@@ -0,0 +1 @@
+test꧁SEP꧂A

test/generate-inputs/expected/meo/meo-targets-expected.txt

@@ -0,0 +1 @@
+C

test/generate-inputs/expected/mincostflow/mincostflow-sources-expected.txt

@@ -0,0 +1 @@
+test_A

test/generate-inputs/expected/mincostflow/mincostflow-targets-expected.txt

@@ -0,0 +1 @@
+C

test/generate-inputs/expected/omicsintegrator1/omicsintegrator1-prizes-expected.txt

@@ -0,0 +1,3 @@
+name	prize
+test_A	2.0
+C	5.7

test/generate-inputs/expected/omicsintegrator2/omicsintegrator2-prizes-expected.txt

@@ -0,0 +1,3 @@
+name	prize
+test_A	2.0
+C	5.7

test/generate-inputs/expected/pathlinker/pathlinker-nodetypes-expected.txt

@@ -0,0 +1,3 @@
+#Node	Node type
+test_A	source
+C	target

test/generate-inputs/expected/responsenet/responsenet-sources-expected.txt

@@ -0,0 +1 @@
+test_A

test/generate-inputs/expected/responsenet/responsenet-targets-expected.txt

@@ -0,0 +1 @@
+C

test/generate-inputs/expected/rwr/rwr-nodes-expected.txt

@@ -0,0 +1,2 @@
+test_A
+C

test/generate-inputs/expected/strwr/strwr-sources-expected.txt

@@ -0,0 +1 @@
+test_A

test/generate-inputs/expected/strwr/strwr-targets-expected.txt

@@ -0,0 +1 @@
+C

test/generate-inputs/test_generate_inputs.py

@@ -1,36 +1,24 @@
 import filecmp
 import os
+import shutil
 from pathlib import Path
 
 import yaml
 
-from spras import runner
-
-OUTDIR = "test/generate-inputs/output/"
-EXPDIR = "test/generate-inputs/expected/"
-
-algo_exp_file = {
-    'mincostflow': 'edges',
-    'meo': 'edges',
-    'omicsintegrator1': 'edges',
-    'omicsintegrator2': 'edges',
-    'domino': 'network',
-    'pathlinker': 'network',
-    'allpairs': 'network',
-    'bowtiebuilder': 'edges',
-    'strwr': 'network',
-    'rwr': 'network',
-    'responsenet': 'edges'
-}
+from spras.runner import algorithms, get_required_inputs, merge_input, prepare_inputs
 
+OUTDIR = Path("test", "generate-inputs", "output")
+EXPDIR = Path("test", "generate-inputs", "expected")
 
 class TestGenerateInputs:
     @classmethod
     def setup_class(cls):
         """
         Create the expected output directory
         """
-        Path(OUTDIR).mkdir(parents=True, exist_ok=True)
+        if OUTDIR.exists():
+            shutil.rmtree(OUTDIR)
+        OUTDIR.mkdir(parents=True, exist_ok=True)
 
     def test_prepare_inputs_networks(self):
         config_loc = os.path.join("test", "generate-inputs", "inputs", "test_config.yaml")
@@ -40,16 +28,15 @@ def test_prepare_inputs_networks(self):
         test_file = "test/generate-inputs/output/test_pickled_dataset.pkl"
 
         test_dataset = next((ds for ds in config["datasets"] if ds["label"] == "test_data"), None)
-        runner.merge_input(test_dataset, test_file)
-
-        for algo in algo_exp_file.keys():
-            inputs = runner.get_required_inputs(algo)
-            filename_map = {input_str: os.path.join("test", "generate-inputs", "output", f"{algo}-{input_str}.txt")
-                            for input_str in inputs}
-            runner.prepare_inputs(algo, test_file, filename_map)
-            exp_file_name = algo_exp_file[algo]
-            assert filecmp.cmp(OUTDIR + f"{algo}-{exp_file_name}.txt", EXPDIR + f"{algo}-{exp_file_name}-expected.txt",
-                               shallow=False)
-
-            for file in filename_map.values():
-                assert Path(file).exists()
+        merge_input(test_dataset, test_file)
+
+        for algo in algorithms.keys():
+            inputs = get_required_inputs(algo)
+            (OUTDIR / algo).mkdir(exist_ok=True)
+            filename_map = {input_str: str(OUTDIR / algo / f"{algo}-{input_str}.txt") for input_str in inputs}
+            prepare_inputs(algo, test_file, filename_map)
+            required_inputs = algorithms[algo].required_inputs
+            for exp_file_name in required_inputs:
+                assert filecmp.cmp(OUTDIR / algo / f"{algo}-{exp_file_name}.txt", EXPDIR / algo / f"{algo}-{exp_file_name}-expected.txt",
+                                   shallow=False), f"{algo} for {exp_file_name}.txt does not match up!"
+