Changeset bda194e3 in sasview

Timestamp:

Aug 21, 2009 10:06:40 AM (15 years ago)

Author:

Jae Cho <jhjcho@…>

Branches:

master, ESS_GUI, ESS_GUI_Docs, ESS_GUI_batch_fitting, ESS_GUI_bumps_abstraction, ESS_GUI_iss1116, ESS_GUI_iss879, ESS_GUI_iss959, ESS_GUI_opencl, ESS_GUI_ordering, ESS_GUI_sync_sascalc, costrafo411, magnetic_scatt, release-4.1.1, release-4.1.2, release-4.2.2, release_4.0.1, ticket-1009, ticket-1094-headless, ticket-1242-2d-resolution, ticket-1243, ticket-1249, ticket885, unittest-saveload

Children:

c1c29b6

Parents:

861a2d3

Message:

This should be the final code for testing 2D 3-axes shape model; comparing 1d to the 2d avg.

File:

• sansmodels/src/sans/models/test/validate_2D_3axes_shape_model.py (modified) (5 diffs)

sansmodels/src/sans/models/test/validate_2D_3axes_shape_model.py

@@ -25 +25 @@
         Averaging the 3-axes 2D scattering intensity give a slightly
         different output than the 1D function  
-        at hight Q (Q>~0.2). This is due(?) to the way the IGOR library 
-        averages, taking only 76 points in alpha, the angle between
+        at hight Q (Q>~0.2). This is due to the way the IGOR library 
+        averages(?), taking only 76 points in alpha, the angle between
         the axis of the ellipsoid and the q vector.
         
@@ -59 +59 @@
         
         npts =points
+        #The average values are very sensitive to npts of phi so npts_alpha should be large enough.
+        npts_alpha =180 #npts of phi
         model = model_class()
         #model.setParam('scale', 1.0)
@@ -84 +86 @@
         output_f.write("<q_average> <2d_average> <1d_average>\n")
             
-        for i_q in range(1, 40):
+        for i_q in range(1, 23):
             q = 0.01*i_q
             sum = 0.0
+            weight = 0.0
             for i_theta in range(npts):
                 theta = math.pi/npts*(i_theta+1)
@@ -92 +95 @@
                 model.setParam(theta_label, theta)
                 
-                for j in range(npts):
-                    model.setParam(phi_label, math.pi * 2.0 / npts * j)
+                for j in range(npts_alpha):
+                    model.setParam(phi_label, math.pi * 2.0 / npts_alpha * j)
                     for k in range(npts):
                         model.setParam(psi_label, math.pi * 2.0 / npts * k)
                         if str(model.run([q, 0])).count("INF")>0:                        
                             print "ERROR", q, theta, math.pi * 2.0 / npts * k
+                            
+                        # sin() is due to having not uniform bin number density wrt the q plane.
                         sum += model.run([q, 0])*math.sin(theta)
-                        #sum += model.run([q, 0])
-            value = sum/npts/(npts)/npts*math.pi/2.0
+                        weight += math.sin(theta) 
+
+            value = sum/weight #*math.pi/2.0
             ana = model.run(q)
-            if q<0.2 and (value-ana)/ana>0.05:
+            if q<0.3 and (value-ana)/ana>0.05:
                 passed = False
             output_f.write("%10g %10g %10g\n" % (q, value, ana))
@@ -115 +121 @@
     validator = Validate2D()
     
-    #te was not passed.
+    #Note: Test one model by one model, otherwise it could crash depending on the memory.
+    
     te_passed =validator(TriaxialEllipsoidModel, points=76)
-    pp_passed = validator(ParallelepipedModel, points=76)
-    ell_passed = validator(EllipticalCylinderModel, points=76)
+    #pp_passed = validator(ParallelepipedModel, points=76)
+    #ell_passed = validator(EllipticalCylinderModel, points=76)
         
     print ""
     print "Model             Passed"
     print "TriaxialEllipsoid         %s" % te_passed
-    print "ParallelepipedModel    %s" % pp_passed
-    print "EllipticalCylinder        %s" % ell_passed
+    #print "ParallelepipedModel    %s" % pp_passed
+    #print "EllipticalCylinder        %s" % ell_passed