From b8ae3de09de7be649c06bb2a91f0e97a03e0ce27 Mon Sep 17 00:00:00 2001
From: Juho Laitala <juho.laitala@gtk.fi>
Date: Fri, 13 Dec 2024 10:36:49 +0200
Subject: [PATCH] added closure checks also when geochemical data that is in
 ppm or ppb

---
 eis_toolkit/utilities/checks/compositional.py | 10 ++++++++--
 tests/utilities/compositional_test.py         | 18 ++++++++++++++++++
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/eis_toolkit/utilities/checks/compositional.py b/eis_toolkit/utilities/checks/compositional.py
index 52e0173d..771f7940 100644
--- a/eis_toolkit/utilities/checks/compositional.py
+++ b/eis_toolkit/utilities/checks/compositional.py
@@ -36,8 +36,14 @@ def check_in_simplex_sample_space(df: pd.DataFrame, tolerance: Number = 0.0001)
     row_sums = df.sum(axis=1)
     closed_to_one = (row_sums - 1).abs() < tolerance
     closed_to_hundred = (row_sums - 100).abs() < tolerance
+    closed_to_million = (row_sums - 1e6).abs() < tolerance
+    closed_to_billion = (row_sums - 1e9).abs() < tolerance
 
-    if not closed_to_one.all() and not closed_to_hundred.all():
-        raise InvalidCompositionException(f"Input data is not closed to 1 or 100 within tolerance of {tolerance}.")
+    is_valid = closed_to_one.all() or closed_to_hundred.all() or closed_to_million.all() or closed_to_billion.all()
+
+    if not is_valid:
+        raise InvalidCompositionException(
+            f"Input data is not closed to 1, 100 (%), 10^6 (ppm) or 10^9 (ppb) within tolerance of {tolerance}."
+        )
 
     return None
diff --git a/tests/utilities/compositional_test.py b/tests/utilities/compositional_test.py
index 1a49453a..d7613f7e 100644
--- a/tests/utilities/compositional_test.py
+++ b/tests/utilities/compositional_test.py
@@ -77,6 +77,9 @@ def test_compositional_data_invalid():
 def test_check_for_simplex_sample_space():
     """Test whether or not a dataframe belongs to a simplex sample space is correctly identified."""
     unit_simplex_df = pd.DataFrame([[0.1, 0.2, 0.3, 0.4], [0.2, 0.3, 0.2, 0.3]])
+    closed_to_hundred_df = unit_simplex_df * 100
+    closed_to_million_df = unit_simplex_df * 1e6
+    closed_to_billion_df = unit_simplex_df * 1e9
     non_simplex_positive_df = pd.DataFrame([1, 2, 3, 4], [5, 6, 7, 8])
     non_positive_df = pd.DataFrame([-1, 2, 3, 4], [1, 2, 3, 4])
 
@@ -91,3 +94,18 @@ def test_check_for_simplex_sample_space():
         check_in_simplex_sample_space(unit_simplex_df)
     except Exception as ex:
         assert False, f"{type(ex)}: {ex}"
+
+    try:
+        check_in_simplex_sample_space(closed_to_hundred_df)
+    except Exception as ex:
+        assert False, f"{type(ex)}: {ex}"
+
+    try:
+        check_in_simplex_sample_space(closed_to_million_df)
+    except Exception as ex:
+        assert False, f"{type(ex)}: {ex}"
+
+    try:
+        check_in_simplex_sample_space(closed_to_billion_df)
+    except Exception as ex:
+        assert False, f"{type(ex)}: {ex}"