Changeset 797 for sans/Dev/trunk/NCNR_User_Procedures/Reduction/SANS

Timestamp:

Apr 8, 2011 12:58:26 PM (12 years ago)

Author:

srkline

Message:

Removed Smear_Model_N_AAO() and reverted back to the non-AAO version. The AAO version is not cursor-aware, and then smeares incorrectly when using cursors. The AAO version can be corrected in the future if necessary. Using the Debye Sphere functions as fit functions need AAO smearing to be useable at all.

File:

• sans/Dev/trunk/NCNR_User_Procedures/Reduction/SANS/WriteQIS.ipf (modified) (1 diff)

sans/Dev/trunk/NCNR_User_Procedures/Reduction/SANS/WriteQIS.ipf

@@ -817 +817 @@
 //*********************
 
-//      // generate my own error wave for I(qx,qy)
+        // generate my own error wave for I(qx,qy)
+        //      //      sw = 0.05*sw            // uniform 5% error? tends to favor the low intensity too strongly
+        // after teh SQRT(),  get rid of the "bad" errors by replacing the NaN, Inf, and zero with V_avg
+        // THIS IS EXTREMEMLY IMPORTANT - if this is not done, there are some "bad" values in the 
+        // error wave (things that are not numbers) - and this wrecks the smeared model fitting.
+        // It appears to have no effect on fitting with the unsmeared model.
 //      Duplicate/O z_val sw
-//      sw = sqrt(z_val)                //assumes Poisson statistics for each cell (counter)
-//      //      sw = 0.05*sw            // uniform 5% error? tends to favor the low intensity too strongly
-//      // get rid of the "bad" errorsby replacing the NaN, Inf, and zero with V_avg
-//      // THIS IS EXTREMEMLY IMPORTANT - if this is not done, there are some "bad" values in the 
-//      // error wave (things that are not numbers) - and this wrecks the smeared model fitting.
-//      // It appears to have no effect on the unsmeared model.
+//      sw = sqrt(z_val)                //assumes Poisson statistics for each cell (counter) - which is wrong, since this is corrected data
+//
 //      WaveStats/Q sw
 //      sw = numtype(sw[p]) == 0 ? sw[p] : V_avg
 //      sw = sw[p] != 0 ? sw[p] : V_avg
-        
-        // now use the properly propagated 2D error
+
+
+//      // April 2011 -- now use the properly propagated 2D error
         Duplicate/O data_err sw
         Redimension/N=(pixelsX*pixelsY) sw