anova

anova.py

@@ -110,49 +110,16 @@ def anova(lst_col_seg, n_level, oa_file, result_file, alpha=0.1):
 
 if __name__ == '__main__':
     # prep data
-    str_sql = """
-    select * from
-    (select distinct idx_scenario, n_max_trial, crit_supplier,
-    firm_pref_request, firm_pref_accept, netw_pref_cust_n,
-    netw_pref_cust_size, cap_limit, diff_new_conn, diff_remove
-    from iiabmdb.with_exp_experiment) as a
-    inner join
-    (
-    select idx_scenario,
-    sum(n_disrupt_s) as n_disrupt_s, sum(n_disrupt_t) as n_disrupt_t from
-    iiabmdb.with_exp_experiment as a
-    inner join
-    (
-    select e_id, count(n_s_disrupt_t) as n_disrupt_s,
-    sum(n_s_disrupt_t) as n_disrupt_t from
-    iiabmdb.with_exp_sample as a
-    inner join
-    (select a.s_id as s_id, count(id) as n_s_disrupt_t from
-    iiabmdb.with_exp_result as a
-    inner join
-    (select distinct s_id from iiabmdb.with_exp_result where ts > 0) as b
-    on a.s_id = b.s_id
-    group by s_id
-    ) as b
-    on a.id = b.s_id
-    group by e_id
-    ) as b
-    on a.id = b.e_id
-    group by idx_scenario) as b
-    on a.idx_scenario = b.idx_scenario;
-
-    """
-    result = pd.read_sql(sql=str_sql,
-                         con=engine)
+    result = pd.read_csv("experiment_result.csv", index_col=None)
     result.drop('idx_scenario', 1, inplace=True)
     df_oa = pd.read_csv("oa_with_exp.csv", index_col=None)
-    result = pd.concat(
+    scenario_result = pd.concat(
         [result.iloc[:, 0:10],
          df_oa.iloc[:, -4:],
          result.iloc[:, -2:]], axis=1)
     result.to_csv('analysis\\experiment_result.csv')
 
-    # 9 factors (X), 4 for error (E), and 2 indicators (Y)
-    the_lst_col_seg = [10, 3, 2]
+    # 10 factors (X), 13 for error (E), and 9 indicators (Y)
+    the_lst_col_seg = [10, 13, 9]
     the_n_level = 3
     anova(the_lst_col_seg, the_n_level, "oa25.txt", result, 0.1)