more cleaning, first two constraints fixed

Author: Jimmy-INL

pysensors/basis/_custom.py

@@ -6,11 +6,7 @@
 
 class Custom(InvertibleBasis, MatrixMixin):
     """
-<<<<<<< HEAD
-    Use a custom transformation to maps input features to
-=======
-    Generate a custom transformation which maps input features to
->>>>>>> 1ccd23fafed0ca39dac6cde204efa228d133df1c
+    Use a custom transformation to map input features to
     custom modes.
 
     Assumes the data has already been centered (to have mean 0).

pysensors/optimizers/_gqr.py

@@ -10,7 +10,7 @@
 
 import pysensors as ps
 from matplotlib.patches import Circle
-
+from pysensors.utils._norm_calc import returnInstance as normCalcReturnInstance
 
 class GQR(QR):
     """
@@ -91,6 +91,7 @@ def fit(self,basis_matrix=None,**optimizer_kws):
             self.all_sensors = None
         if 'constraint_option' in optimizer_kws.keys():
             self.constraint_option = optimizer_kws['constraint_option']
+            self._norm_calc_Instance = normCalcReturnInstance(self,self.constraint_option)
         else:
             self.constraint_option = None
         if 'nx' in optimizer_kws.keys():
@@ -127,34 +128,36 @@ def fit(self,basis_matrix=None,**optimizer_kws):
 
             # Norm of each column
             dlens = np.sqrt(np.sum(np.abs(r) ** 2, axis=0))
-
-            if self.constraint_option == "max_n_const_sensors" :
-                dlens_updated = ps.utils._norm_calc.norm_calc_max_n_const_sensors(self.constrainedIndices,dlens,p,j, self.nConstrainedSensors,self.all_sensorloc,self.n_sensors)
-                i_piv = np.argmax(dlens_updated)
-                dlen = dlens_updated[i_piv]
-            elif self.constraint_option == "exact_n_const_sensors" :
-                dlens_updated = ps.utils._norm_calc.norm_calc_exact_n_const_sensors(self.constrainedIndices,dlens,p,j,self.nConstrainedSensors)
-                i_piv = np.argmax(dlens_updated)
-                dlen = dlens_updated[i_piv]
-            elif self.constraint_option == "predetermined_end":
-                dlens_updated = ps.utils._norm_calc.predetermined_norm_calc(self.constrainedIndices, dlens, p, j, self.nConstrainedSensors, self.n_sensors)
-                i_piv = np.argmax(dlens_updated)
-                dlen = dlens_updated[i_piv]
-            elif self.constraint_option == "radii_constraints":
-
-                if j == 0:
-                    i_piv = np.argmax(dlens)
-                    dlen = dlens[i_piv]
-                    dlens_old = dlens
-                else:
-
-                    dlens_updated = ps.utils._norm_calc.f_radii_constraint(j,dlens,dlens_old,p,self._nx,self._ny,self._r,self.all_sensorloc, self.n_sensors) #( self.radius,self._nx,self._ny,self.all_sensorloc,dlens,p,j)
-                    i_piv = np.argmax(dlens_updated)
-                    dlen = dlens_updated[i_piv]
-                    dlens_old = dlens_updated
-            else:
-                i_piv = np.argmax(dlens)
-                dlen = dlens[i_piv]
+            dlens_updated = self._norm_calc_Instance(self.constrainedIndices, dlens, p, j, self.nConstrainedSensors, all_sensors=self.all_sensors, n_sensors=self.n_sensors, nx=self._nx, ny=self._ny, r=self._r)
+            i_piv = np.argmax(dlens_updated)
+            dlen = dlens_updated[i_piv]
+            # if self.constraint_option == "max_n_const_sensors" :
+            #     dlens_updated = ps.utils._norm_calc.norm_calc_max_n_const_sensors(self.constrainedIndices,dlens,p,j, self.nConstrainedSensors,self.all_sensorloc,self.n_sensors)
+            #     i_piv = np.argmax(dlens_updated)
+            #     dlen = dlens_updated[i_piv]
+            # elif self.constraint_option == "exact_n_const_sensors" :
+            #     dlens_updated = ps.utils._norm_calc.norm_calc_exact_n_const_sensors(self.constrainedIndices,dlens,p,j,self.nConstrainedSensors)
+            #     i_piv = np.argmax(dlens_updated)
+            #     dlen = dlens_updated[i_piv]
+            # elif self.constraint_option == "predetermined_end":
+            #     dlens_updated = ps.utils._norm_calc.predetermined_norm_calc(self.constrainedIndices, dlens, p, j, self.nConstrainedSensors, self.n_sensors)
+            #     i_piv = np.argmax(dlens_updated)
+            #     dlen = dlens_updated[i_piv]
+            # elif self.constraint_option == "radii_constraints":
+
+            #     if j == 0:
+            #         i_piv = np.argmax(dlens)
+            #         dlen = dlens[i_piv]
+            #         dlens_old = dlens
+            #     else:
+
+            #         dlens_updated = ps.utils._norm_calc.f_radii_constraint(j,dlens,dlens_old,p,self._nx,self._ny,self._r,self.all_sensorloc, self.n_sensors) #( self.radius,self._nx,self._ny,self.all_sensorloc,dlens,p,j)
+            #         i_piv = np.argmax(dlens_updated)
+            #         dlen = dlens_updated[i_piv]
+            #         dlens_old = dlens_updated
+            # else:
+            #     i_piv = np.argmax(dlens)
+            #     dlen = dlens[i_piv]
 
             # Choose pivot
             # i_piv = np.argmax(dlens_updated)

pysensors/utils/_norm_calc.py

@@ -4,9 +4,9 @@
 
 import numpy as np
 
-def norm_calc_exact_n_const_sensors(lin_idx, dlens, piv, j, n_const_sensors): ##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 norm_calc_exact_n_const_sensors(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
     """
     Function for mapping constrained sensor locations with the QR procedure.
 
@@ -27,18 +27,11 @@ def norm_calc_exact_n_const_sensors(lin_idx, dlens, piv, j, n_const_sensors): ##
         -------
         dlens : np.darray, shape [Variable based on j] with constraints mapped into it.
     """
-    if j < n_const_sensors: # force sensors into constraint region
-        #idx = np.arange(dlens.shape[0])
-        #dlens[np.delete(idx, lin_idx)] = 0
-
-        didx = np.isin(piv[j:],lin_idx,invert=True)
-        dlens[didx] = 0
-    else:
-        didx = np.isin(piv[j:],lin_idx,invert=False)
-        dlens[didx] = 0
+    didx = np.isin(piv[j:],lin_idx,invert=j<n_const_sensors)
+    dlens[didx] = 0
     return dlens
 
-def norm_calc_max_n_const_sensors(lin_idx, dlens, piv, j, const_sensors,all_sensors,n_sensors): ##Optimal sensor placement with constraints (will place sensors in  the order of QR)
+def norm_calc_max_n_const_sensors(lin_idx, dlens, piv, j, n_const_sensors, **kwargs):
     """
     Function for mapping constrained sensor locations with the QR procedure (Optimally).
 
@@ -63,21 +56,29 @@ def norm_calc_max_n_const_sensors(lin_idx, dlens, piv, j, const_sensors,all_sens
         -------
         dlens : np.darray, shape [Variable based on j] with constraints mapped into it.
     """
+    if 'all_sensors' in kwargs.keys():
+        all_sensors = kwargs['all_sensors']
+    else:
+        all_sensors = []
+    if 'n_sensors' in kwargs.keys():
+        n_sensors = kwargs['n_sensors']
+    else:
+        n_sensors = len(all_sensors)
     counter = 0
     mask = np.isin(all_sensors,lin_idx,invert=False)
     const_idx = all_sensors[mask]
-    updated_lin_idx = const_idx[const_sensors:]
+    updated_lin_idx = const_idx[n_const_sensors:]
     for i in range(n_sensors):
         if np.isin(all_sensors[i],lin_idx,invert=False):
             counter += 1
-            if counter < const_sensors:
+            if counter < n_const_sensors:
                 dlens = dlens
             else:
                 didx = np.isin(piv[j:],updated_lin_idx,invert=False)
                 dlens[didx] = 0
     return dlens
 
-def predetermined_norm_calc(lin_idx, dlens, piv, j, n_const_sensors, n_sensors):
+def predetermined_norm_calc(lin_idx, dlens, piv, j, n_const_sensors, **kwargs):
     """
     Function for mapping constrained sensor locations with the QR procedure.
 
@@ -98,15 +99,13 @@ def predetermined_norm_calc(lin_idx, dlens, piv, j, n_const_sensors, n_sensors):
         -------
         dlens : np.darray, shape [Variable based on j] with constraints mapped into it.
     """
-    if (n_sensors - n_const_sensors) <= j <= n_sensors: # force sensors into constraint region
-        #idx = np.arange(dlens.shape[0])
-        #dlens[np.delete(idx, lin_idx)] = 0
-
-        didx = np.isin(piv[j:],lin_idx,invert=True)
-        dlens[didx] = 0
+    if 'n_sensors' in kwargs.keys():
+        n_sensors = kwargs['n_sensors']
     else:
-        didx = np.isin(piv[j:],lin_idx,invert=False)
-        dlens[didx] = 0
+        raise ValueError ('total number of sensors is not given!')
+
+    didx = np.isin(piv[j:],lin_idx,invert=(n_sensors - n_const_sensors) <= j <= n_sensors)
+    dlens[didx] = 0
     return dlens
 
 def f_radii_constraint(j,dlens,dlens_old,piv,nx,ny,r, all_sensors, n_sensors):
@@ -122,7 +121,7 @@ def f_radii_constraint(j,dlens,dlens_old,piv,nx,ny,r, all_sensors, n_sensors):
         result_list = [x + (j-1) for x in result_list]
         result_array = np.array(result_list)
         print(result_array)
-       
+
         idx_constrained1 = get_constraind_sensors_indices_radii(j,piv,r, nx,ny, all_sensors)
         t = np.concatenate((idx_constrained1,result_array), axis = 0)
         didx = np.isin(piv[j:],t,invert= False)
@@ -166,4 +165,20 @@ def get_constraind_sensors_indices_radii(j,piv,r, nx,ny, all_sensors):
     else:
         idx_constrained = np.ravel_multi_index(constrained_sensors_tuple, (nx,ny))
     return idx_constrained
-    
+
+
+__norm_calc_type = {}
+__norm_calc_type['exact_n_const_sensors'] = norm_calc_exact_n_const_sensors
+__norm_calc_type['max_n_const_sensors'] = norm_calc_max_n_const_sensors
+__norm_calc_type['predetermined_norm_calc'] = predetermined_norm_calc
+
+def returnInstance(cls, name):
+  """
+    Method designed to return class instance:
+    @ In, cls, class type
+    @ In, name, string, name of class
+    @ Out, __crossovers[name], instance of class
+  """
+  if name not in __norm_calc_type:
+    cls.raiseAnError (IOError, "{} NOT IMPLEMENTED!!!!!".format(name))
+  return __norm_calc_type[name]