#313 #287 #281 use CI to test very mini build (only fastest ETL); tes…

…t reporting of semmed exclusion
RTXteam · Jul 18, 2023 · acdf781 · acdf781
1 parent 1489445
commit acdf781
Show file tree

Hide file tree

Showing 2 changed files with 25 additions and 0 deletions.
diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml
@@ -35,3 +35,10 @@ jobs:
         bash -x ./build-kg2-snakemake.sh all -R_Merge -n nodes ci
         bash -x ./build-kg2-snakemake.sh all -R_Finish -n nodes ci
         bash -x ./build-kg2-snakemake.sh all -F -n nodes ci
+    - name: Test Building One File
+      run: |
+        cd /home/runner/work/RTX-KG2/RTX-KG2/RTX-KG2
+        bash -x ./extract-mirbase.sh ~/kg2-build/miRNA.dat
+        ~/kg2-venv/bin/python3 -u mirbase_dat_to_kg_json.py ~/kg2-build/miRNA.dat ~/kg2-build/kg2-mirbase.json --test
+        ~/kg2-venv/bin/python3 -u report_stats_on_json_kg.py ~/kg2-build/kg2-mirbase.json ~/kg2-build/kg2-mirbase-report.json
+        cat ~/kg2-build/kg2-mirbase-report.json
diff --git a/report_stats_on_json_kg.py b/report_stats_on_json_kg.py
@@ -163,6 +163,24 @@ def get_deprecated_nodes(nodes: list):
 
     return deprecated_nodes
 
+
+def get_excluded_nodes(nodes: list):
+    excluded_nodes = dict()
+    provided_by_label = 'provided_by'
+    excluded_label = 'domain_range_exclusion'
+
+    for node in nodes:
+        source = node[provided_by_label][0]
+        excluded = node[excluded_label]
+
+        if excluded:
+            if source not in excluded_nodes:
+                excluded_nodes[source] = 0
+            excluded_nodes[source] += 1
+
+    return excluded_nodes
+
+
 def count_orphan_nodes(nodes: list, edges: list):
     orphan_nodes = dict()
     provided_by_label = 'provided_by'