Bo klsm (#16)

* klsm_code added

* fix issues with klsm

* debugging k

* still debugging

* trying to fix bounds

* still fixing bounds

* fixing possible fixed_features

* fixing possible fixed_features again

* fixing possible fixed_features still

* modified qlsm to evalyuate based on klsm

* fixed issue with design printing

* refactor code, bug fix

* fixed logic for k_level

* klsm without debug code

endure.toml

@@ -153,33 +153,37 @@ drop_last = true
 [job.BayesianOptimization]
 # -----------------------------------------------------------------------------
 num_iterations = 15
-num_restarts = 20
+num_restarts = 2 # TODO set it to 2
 # value of raw_samples determines how many initial random samples are taken from the search space before starting the optimization process
-raw_samples = 30
+raw_samples = 3 # TODO set it to 3
 initial_samples = 20
+# for a true KLSM calculation - set num_k_values to the same value as max_levels. This is only consequential for the KLSM model
+# This works in the following way:
+# suppose num_k_values = 4 and max_levels = 20
+# Then every layer till the 4th layer will have custom k values but the (num_k_values + 1) layer to the (max_levels) layer will only
+# have a k value equal to 1
+num_k_values = 4
 
 # This is the q value used in BoTorch Acquisition functions.
 # if it is set to a value above 1 sequential processing will stop in acquisition function and batch processing will start
 # note that for batch processing tensor shape will change and will require modification of code.
 # TODO: Add code to handle batch
 batch_size = 1
-max_levels = 16
-
 # Acquisition function options
 # [ExpectedImprovement, UpperConfidenceBound, qExpectedImprovement]
 acquisition_function = "ExpectedImprovement"
 beta_value = 0.3
 # model_type can take values - "Classic", "QFixed", "YZHybrid", "KHybrid"
-model_type = "QFixed"
+model_type = "KHybrid"
 
 [job.BayesianOptimization.database]
 # This will take value 0 and 1 where 1 means write each cost and run details into the MySqlLite database
 # and 0 means run details are not stored in the database
 write_to_db = 1
 # by default the databases directory will be created inside the data director. To change this, you need to change ["io"]["data_dir"]
-db_path = "databases"
+db_path = "yz_databases"
 # This must be a .db file for code to function. It will create a sqllite database
-db_name = "db_cost.db"
+db_name = "yz_db_cost.db"
 
 [job.BayesianOptimization.system]
 E = 1024

endure/util/db_log.py

@@ -33,6 +33,10 @@ def initialize_database(db_path='cost_log.db'):
             Q INTEGER,
             Y INTEGER,
             Z INTEGER,
+            k1 REAL, k2 REAL, k3 REAL, k4 REAL, k5 REAL,
+            k6 REAL, k7 REAL, k8 REAL, k9 REAL, k10 REAL,
+            k11 REAL, k12 REAL, k13 REAL, k14 REAL, k15 REAL,
+            k16 REAL, k17 REAL, k18 REAL, k19 REAL, k20 REAL,
             cost REAL,
             FOREIGN KEY (run_id) REFERENCES runs(run_id)
         );''')
@@ -56,6 +60,17 @@ def initialize_database(db_path='cost_log.db'):
             bayesian_Z INTEGER,
             FOREIGN KEY (run_id) REFERENCES runs(run_id)
         );''')
+    cursor.execute('''
+        CREATE TABLE IF NOT EXISTS run_details_k_values (
+            idx INTEGER PRIMARY KEY AUTOINCREMENT,
+            run_id INTEGER,
+            k1 REAL, k2 REAL, k3 REAL, k4 REAL, k5 REAL,
+            k6 REAL, k7 REAL, k8 REAL, k9 REAL, k10 REAL,
+            k11 REAL, k12 REAL, k13 REAL, k14 REAL, k15 REAL,
+            k16 REAL, k17 REAL, k18 REAL, k19 REAL, k20 REAL,
+            FOREIGN KEY (run_id) REFERENCES runs(run_id)
+        );''')
+
     connector.commit()
     return connector
 
@@ -76,9 +91,12 @@ def log_new_run(connector, system, workload, iterations, sample, acqf):
 def log_design_cost(connector, run_id, design, cost):
     cursor = connector.cursor()
     policy = design.policy
-    cursor.execute('INSERT INTO design_costs (run_id, bits_per_element, size_ratio, policy, Q, Y, Z, cost) '
-                   'VALUES (?, ?, ?, ?, ?, ?, ?, ?)', (run_id, design.h, design.T, policy.name, design.Q, design.Y,
-                                                       design.Z, cost))
+    k_values = design.K + [None] * (20 - len(design.K))  # TODO replace this with the max_levels
+    sql_command = ('INSERT INTO design_costs (run_id, bits_per_element, size_ratio, policy, Q, Y, Z, cost, ' +
+                   ', '.join([f'k{i+1}' for i in range(20)]) + ') ' + 'VALUES (?, ?, ?, ?, ?, ?, ?, ?'', ' +
+                   ', '.join(['?']*20) + ')')
+    cursor.execute(sql_command, (run_id, design.h, design.T, policy.name, design.Q, design.Y, design.Z, cost) +
+                   tuple(k_values))
     connector.commit()
 
 
@@ -87,18 +105,20 @@ def log_run_details(connector, run_id, duration, analytical_cost, bayesian_cost,
     analytical_policy = analytical_design.policy
     print(analytical_policy)
     bayesian_policy = bayesian_design.policy
-    if bayesian_policy == 0.0:
-        bo_policy = Policy.Leveling
-    else:
-        bo_policy = Policy.Tiering
     cursor.execute('''
         INSERT INTO run_details (run_id, duration_secs, analytical_cost, bayesian_cost, analytical_h, analytical_T, 
         analytical_policy, analytical_Q, analytical_Y, analytical_Z, bayesian_h, bayesian_T, bayesian_policy, bayesian_Q, 
         bayesian_Y, bayesian_Z) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)''',
                    (run_id, duration, analytical_cost, bayesian_cost, analytical_design.h, analytical_design.T,
                     analytical_policy.name, analytical_design.Q, analytical_design.Y, analytical_design.Z,
-                    bayesian_design.h, bayesian_design.T, bo_policy.name, bayesian_design.Q,
+                    bayesian_design.h, bayesian_design.T, bayesian_policy.name, bayesian_design.Q,
                     bayesian_design.Y, bayesian_design.Z))
+    if bayesian_policy == Policy.KHybrid:
+        k_values = bayesian_design.K + [None] * (20 - len(bayesian_design.K))
+        cursor.execute('''
+            INSERT INTO run_details_k_values (run_id, k1, k2, k3, k4, k5, k6, k7, k8, k9, k10, k11, k12, k13, k14, k15, 
+            k16, k17, k18, k19, k20) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)''',
+                       (run_id,) + tuple(k_values))
     connector.commit()