From b829bf6f5a26edaa57acde0c4a21e2c24d695e87 Mon Sep 17 00:00:00 2001
From: Frederick Muriuki Muriithi
Date: Tue, 19 Oct 2021 09:16:38 +0300
Subject: Implement `dictify_by_samples`

Issue:
https://github.com/genenetwork/gn-gemtext-threads/blob/main/topics/gn1-migration-to-gn2/partial-correlations.gmi

* gn3/partial_correlations.py: implement `dictify_by_samples` function
* tests/unit/test_partial_correlations.py: implement tests for
  `dictify_by_samples` function

  Implement the `dictify_by_samples` function as a partial migration of the
  `web.webqtl.correlation.correlationFunction.fixStrains` function from GN1.
---
 gn3/partial_correlations.py             | 16 +++++++++++++
 tests/unit/test_partial_correlations.py | 42 ++++++++++++++++++++++++++++++++-
 2 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/gn3/partial_correlations.py b/gn3/partial_correlations.py
index 99521c6..4db4807 100644
--- a/gn3/partial_correlations.py
+++ b/gn3/partial_correlations.py
@@ -44,3 +44,19 @@ def control_samples(controls: Sequence[dict], sampleslist: Sequence[str]):
         ),
         [__process_control__(trait_data) for trait_data in controls],
         (tuple(), tuple(), tuple(), tuple()))
+
+def dictify_by_samples(samples_vals_vars: Sequence[Sequence]) -> dict:
+    """
+    Build a sequence of dictionaries from a sequence of separate sequences of
+    samples, values and variances.
+
+    This is a partial migration of
+    `web.webqtl.correlation.correlationFunction.fixStrains` function in GN1.
+    This implementation extracts code that will find common use, and that will
+    find use in more than one place.
+    """
+    return tuple(
+        {
+            sample: {"sample_name": sample, "value": val, "variance": var}
+            for sample, val, var in zip(*trait_line)
+        } for trait_line in zip(*(samples_vals_vars[0:3])))
diff --git a/tests/unit/test_partial_correlations.py b/tests/unit/test_partial_correlations.py
index 0083ef7..6302f74 100644
--- a/tests/unit/test_partial_correlations.py
+++ b/tests/unit/test_partial_correlations.py
@@ -1,7 +1,7 @@
 """Module contains tests for gn3.partial_correlations"""
 
 from unittest import TestCase
-from gn3.partial_correlations import control_samples
+from gn3.partial_correlations import control_samples, dictify_by_samples
 
 sampleslist = ["B6cC3-1", "BXD1", "BXD12", "BXD16", "BXD19", "BXD2"]
 control_traits = (
@@ -85,3 +85,43 @@ class TestPartialCorrelations(TestCase):
              ((None, None, None, None, None, None), (None, None, None, None),
               (None, None, None)),
              (6, 4, 3)))
+
+    def test_dictify_by_samples(self):
+        """
+        Given:
+            a sequence of sequences with sample names, values and variances, as
+            in the output of `gn3.partial_correlations.control_samples` or
+            the output of `gn3.db.traits.export_informative`
+        When:
+            the sequence is passed as an argument into the
+            `gn3.partial_correlations.dictify_by_sample`
+        Then:
+            return a sequence of dicts with keys being the values of the sample
+            names, and each of who's values being sub-dicts with the keys
+            'sample_name', 'value' and 'variance' whose values correspond to the
+            values passed in.
+        """
+        self.assertEqual(
+            dictify_by_samples(
+                ((("B6cC3-1", "BXD1", "BXD12", "BXD16", "BXD19", "BXD2"),
+                  ("BXD12", "BXD16", "BXD19", "BXD2"),
+                  ("B6cC3-1", "BXD1", "BXD2")),
+                 ((7.51879, 7.77141, 8.39265, 8.17443, 8.30401, 7.80944),
+                  (8.39265, 8.17443, 8.30401, 7.80944),
+                  (7.51879, 7.77141, 7.80944)),
+                 ((None, None, None, None, None, None), (None, None, None, None),
+                  (None, None, None)),
+                 (6, 4, 3))),
+            ({"B6cC3-1": {"sample_name": "B6cC3-1", "value": 7.51879, "variance": None},
+              "BXD1": {"sample_name": "BXD1", "value": 7.77141, "variance": None},
+              "BXD12": {"sample_name": "BXD12", "value": 8.39265, "variance": None},
+              "BXD16": {"sample_name": "BXD16", "value": 8.17443, "variance": None},
+              "BXD19": {"sample_name": "BXD19", "value": 8.30401, "variance": None},
+              "BXD2": {"sample_name": "BXD2", "value": 7.80944, "variance": None}},
+             {"BXD12": {"sample_name": "BXD12", "value": 8.39265, "variance": None},
+              "BXD16": {"sample_name": "BXD16", "value": 8.17443, "variance": None},
+              "BXD19": {"sample_name": "BXD19", "value": 8.30401, "variance": None},
+              "BXD2": {"sample_name": "BXD2", "value": 7.80944, "variance": None}},
+             {"B6cC3-1": {"sample_name": "B6cC3-1", "value": 7.51879, "variance": None},
+              "BXD1": {"sample_name": "BXD1", "value": 7.77141, "variance": None},
+              "BXD2": {"sample_name": "BXD2", "value":  7.80944, "variance": None}}))
-- 
cgit v1.2.3