--- a
+++ b/benchmark/disease2icd.py
@@ -0,0 +1,58 @@
+'''
+input:  
+
+	"data/drugbank_trial_conditions.csv"
+
+process:
+	disease maps to icd? code 
+
+output: 
+
+	"data/disease2icd.pkl" 
+
+
+
+'''
+
+
+
+
+
+
+import csv, pickle 
+from collections import defaultdict 
+def disease2icd_func():
+	file = "data/drugbank_trial_conditions.csv"
+	with open(file, 'r') as csvfile:
+		reader = list(csv.reader(csvfile, delimiter = ','))[1:]
+	disease2icdcode = defaultdict(set)
+	disease2icdcode2 = dict()
+	for row in reader:
+		diseasename1 = row[2].lower()
+		diseasename2 = row[6].lower()
+		icd10code = row[8]
+		if icd10code.strip() == '':
+			continue 
+		disease2icdcode[diseasename1].add(icd10code)
+		disease2icdcode[diseasename2].add(icd10code)
+	for disease, icdcode in disease2icdcode.items():
+		assert len(icdcode)==1 
+		disease2icdcode2[disease] = list(icdcode)[0]
+	return disease2icdcode2 
+
+
+### disease -> icd code
+if __name__ == "__main__":
+	disease2icdcode = disease2icd_func()
+	pickle.dump(disease2icdcode, open("data/disease2icd.pkl", 'wb')) 
+	for disease, icd in disease2icdcode.items():
+		if len(disease.split())==1:
+			print(disease, icd)
+
+
+
+
+
+
+
+