Adding max_n and predetermined tests to test_optimizers.py, deleting comments from _gqr.py,spatially_constrained_qr.ipynb and _norm_calc-py,adding kwargs documentation to _custom.py to address Josh comments

Author: niharika2999

examples/spatially_constrained_qr.ipynb

@@ -27,6 +27,9 @@ class Custom(InvertibleBasis, MatrixMixin):
     """
 
     def __init__(self, U, n_basis_modes=10, **kwargs):
+        '''
+            kwargs : Not defined but added to remain consistent with prior basis functions. 
+        '''
         if isinstance(n_basis_modes, int) and n_basis_modes > 0:
             super(Custom, self).__init__()#n_components=n_basis_modes, **kwargs
             self._n_basis_modes = n_basis_modes

pysensors/basis/_custom.py

@@ -26,6 +26,10 @@ class GQR(QR):
         "Data-driven sparse sensor placement for reconstruction:
         Demonstrating the benefits of exploiting known patterns."
         IEEE Control Systems Magazine 38.3 (2018): 63-86.
+        
+        Niharika Karnik, Mohammad G. Abdo, Carlos E. Estrada Perez, Jun Soo Yoo, Joshua J. Cogliati, Richard S. Skifton, 
+        Pattrick Calderoni, Steven L. Brunton, and Krithika Manohar. 
+        Optimal Sensor Placement with Adaptive Constraints for Nuclear Digital Twins. 2023. arXiv: 2306 . 13637 [math.OC].
 
     @ authors: Niharika Karnik (@nkarnik2999), Mohammad Abdo (@Jimmy-INL), and Krithika Manohar (@kmanohar)
     """
@@ -76,13 +80,6 @@ def fit(self,basis_matrix,**optimizer_kws):
         n_features, n_samples = basis_matrix.shape  # We transpose basis_matrix below
         max_const_sensors = len(self.idx_constrained) # Maximum number of sensors allowed in the constrained region
 
-        ## Assertions and checks:
-        # if self.n_sensors > n_features - max_const_sensors + self.nConstrainedSensors:
-        #     raise IOError ("n_sensors cannot be larger than n_features - all possible locations in the constrained area + allowed constrained sensors")
-        # if self.n_sensors > n_samples + self.nConstrainedSensors: ## Handling zero constraint?
-        #     raise IOError ("Currently n_sensors should be less than min(number of samples, number of modes) + number of constrained sensors,\
-        #                    got: n_sensors = {}, n_samples + const_sensors = {} + {} = {}".format(self.n_sensors,n_samples,self.nConstrainedSensors,n_samples+self.nConstrainedSensors))
-
         # Initialize helper variables
         R = basis_matrix.conj().T.copy()
         p = np.arange(n_features)

pysensors/optimizers/_gqr.py

@@ -7,9 +7,7 @@
 def unconstrained(lin_idx, dlens, piv, j, n_const_sensors, **kwargs):
     return dlens
 
-def exact_n(lin_idx, dlens, piv, j, n_const_sensors, **kwargs): ##Will first force sensors into constrained region
-    # num_sensors should be fixed for each custom constraint (for now)
-    # num_sensors must be <= size of constraint region
+def exact_n(lin_idx, dlens, piv, j, n_const_sensors, **kwargs): 
     """
     Function for mapping constrained sensor locations with the QR procedure.
 

pysensors/utils/_norm_calc.py

@@ -107,12 +107,50 @@ def test_gqr_exact_constrainted_case1(data_random):
 
     # Get ranked sensors from GQR
     sensors_GQR = GQR().fit(x.T, idx_constrained=forbidden_sensors,n_sensors=total_sensors,n_const_sensors=exact_n_const_sensors, constraint_option='exact_n_const_sensors').get_sensors()[:total_sensors]
+    assert sensors_GQR.intersection(forbidden_sensors) == 3
+   
 
-    # try to compare these using the validation metrics
+def test_gqr_max_constrained_case1(data_random):
+    ## In this case we want to place a total of 10 sensors
+    # with a constrained region that is allowed to have a maximum of 3 sensors
+    # but 4 of the first 10 are in the constrained region
+    x = data_random
+    # unconstrained sensors (optimal)
+    sensors_QR = QR().fit(x.T).get_sensors()
+    # exact number of sensors allowed in the constrained region
+    total_sensors = 10
+    max_n_const_sensors = 3
+    forbidden_sensors = [8,5,2,6]
+    totally_forbidden_sensors = [x for x in forbidden_sensors if x in sensors_QR][:max_n_const_sensors]
+    totally_forbidden_sensors = [y for y in forbidden_sensors if y not in totally_forbidden_sensors]
+    costs = np.zeros(x.shape[1])
+    costs[totally_forbidden_sensors] = 100
+    # Get ranked sensors
+    sensors = CCQR(sensor_costs=costs).fit(x.T).get_sensors()[:total_sensors]
 
-## TODO
-def test_gqr_max_constrained():
-    pass
+    # Forbidden sensors should not be included
+    chosen_sensors = set(sensors[: (x.shape[1] - len(totally_forbidden_sensors))])
+    assert chosen_sensors.isdisjoint(set(totally_forbidden_sensors))
+
+
+    # Get ranked sensors from GQR
+    sensors_GQR = GQR().fit(x.T, idx_constrained=forbidden_sensors,n_sensors=total_sensors,n_const_sensors=max_n_const_sensors, constraint_option='max_n_const_sensors').get_sensors()[:total_sensors]
+    assert sensors_GQR.intersection(forbidden_sensors) == 3
+    
+def test_gqr_predetermined_case1(data_random):
+    ## In this case we want to place a total of 10 sensors
+    # 2 of the sensors are predetermined by the user
+    x = data_random
+    # unconstrained sensors (optimal)
+    sensors_QR = QR().fit(x.T).get_sensors()
+    # Predtermined sensors
+    total_sensors = 10
+    n_sensors_pre = 2
+    predetermined_sensors = [8,5]
+
+    # Predetermined sensors shopuld be included
+    # Get ranked sensors from GQR
+    sensors_GQR = GQR().fit(x.T, idx_constrained=predetermined_sensors,n_sensors=total_sensors,n_const_sensors=n_sensors_pre, constraint_option='predetermined').get_sensors()[:total_sensors]
+    assert sensors_GQR.intersection(predetermined_sensors) == 2
+    
 
-def test_gqr_radii_constrained():
-    pass