From 1544776b072d7240773cf14d423078841e4c1a07 Mon Sep 17 00:00:00 2001
From: Frederick Muriuki Muriithi
Date: Mon, 18 Oct 2021 14:14:04 +0300
Subject: Implement `control_samples` function as is in GN1

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

* gn3/partial_correlations.py: Implement `control_samples` function
* tests/unit/test_partial_correlations.py: add tests for `control_samples`
  function

  Implement the function `control_samples` and make it mostly bug-compatible
  with the `web/webqtl/correlation/correlationFunction.controlStrain` function
  in GN1.

  This implementation in GN3 does not do any calls to the database. It will
  rely on other functions to provide the data from the database to it.
---
 gn3/partial_correlations.py | 38 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 38 insertions(+)

(limited to 'gn3/partial_correlations.py')

diff --git a/gn3/partial_correlations.py b/gn3/partial_correlations.py
index df390ed..99521c6 100644
--- a/gn3/partial_correlations.py
+++ b/gn3/partial_correlations.py
@@ -5,4 +5,42 @@ It is an attempt to migrate over the partial correlations feature from
 GeneNetwork1.
 """
 
+from typing import Sequence
 from functools import reduce
+
+def control_samples(controls: Sequence[dict], sampleslist: Sequence[str]):
+    """
+    Fetches data for the control traits.
+
+    This migrates `web/webqtl/correlation/correlationFunction.controlStrain` in
+    GN1, with a few modifications to the arguments passed in.
+
+    PARAMETERS:
+    controls: A map of sample names to trait data. Equivalent to the `cvals`
+        value in the corresponding source function in GN1.
+    sampleslist: A list of samples. Equivalent to `strainlst` in the
+        corresponding source function in GN1
+    """
+    def __process_control__(trait_data):
+        def __process_sample__(acc, sample):
+            if sample in trait_data["data"].keys():
+                sample_item = trait_data["data"][sample]
+                val = sample_item["value"]
+                if val is not None:
+                    return (
+                        acc[0] + (sample,),
+                        acc[1] + (val,),
+                        acc[2] + (sample_item["variance"],))
+            return acc
+        return reduce(
+            __process_sample__, sampleslist, (tuple(), tuple(), tuple()))
+
+    return reduce(
+        lambda acc, item: (
+            acc[0] + (item[0],),
+            acc[1] + (item[1],),
+            acc[2] + (item[2],),
+            acc[3] + (len(item[0]),),
+        ),
+        [__process_control__(trait_data) for trait_data in controls],
+        (tuple(), tuple(), tuple(), tuple()))
-- 
cgit v1.2.3