Reorganized folder system. added an assorted section to help keep track of useful sql etc.

assorted/examining the db.sql

@@ -0,0 +1,44 @@
+SELECT why_stopped FROM ctgov.studies
+WHERE why_stopped IS NOT NULL
+LIMIT 100;
+
+SELECT study_type, count(*) from ctgov.studies
+group by study_type;
+
+SELECT is_fda_regulated_drug, count(*) from ctgov.studies
+GROUP BY is_fda_regulated_drug;
+
+
+
+/*
+Note that there is a decent number of trials that have expanded access
+*/
+SELECT 
+	study_type
+	, phase
+	, has_expanded_access
+	, has_dmc
+	, count(*) 
+FROM ctgov.studies
+WHERE 
+	is_fda_regulated_drug is true
+	AND
+	study_type = 'Interventional'
+	AND 
+	start_date > date('2007-01-01')
+group by study_type, phase, has_expanded_access, has_dmc;
+
+
+
+
+/*
+ Find different mesh terms as assigned by clinicaltrials.gov
+ */
+select * from ctgov.browse_conditions
+order by nct_id desc,mesh_type
+limit 200;
+
+select * from ctgov.browse_interventions
+order by nct_id desc
+limit 200;
+

classifications/classify_terminations.py

@@ -38,11 +38,11 @@ if __name__ == "__main__":
 
     X = vectorizer.fit_transform(data.cleaned)
 
-    kmeans = KMeans(n_clusters=3, random_state=11021585)
+    kmeans = KMeans(n_clusters=10, random_state=11021585)
     kmeans.fit(X)
 
     data["cluster"] = kmeans.labels_
 
-    print(data.groupby(["cluster"]).count())
+    print(data.groupby(["cluster"])["cleaned"].count())