Changeset feb21d8 in sasview for prview

Timestamp:

Jul 8, 2008 11:37:00 AM (16 years ago)

Author:

Mathieu Doucet <doucetm@…>

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:

5615a7b

Parents:

6e0f53a

Message:

Added various scaling choices

Location:

prview

Files:

• local_config.py (modified) (1 diff)
• perspectives/pr/pr.py (modified) (8 diffs)
• perspectives/pr/requirements.txt (modified) (1 diff)
• release_notes.txt (modified) (1 diff)

prview/local_config.py

@@ -6 +6 @@
 # Version of the application
 __appname__ = "PrView"
-__version__ = '0.2.1'
+__version__ = '0.2.2'
 __download_page__ = 'http://danse.chem.utk.edu'
+__update_URL__ = 'http://danse.chem.utk.edu/prview_version.txt'
 
 

prview/perspectives/pr/pr.py

@@ -60 +60 @@
         self.invertor    = None
         self.pr          = None
-        self.pr_estimate = None
+        # Copy of the last result in case we need to display it.
+        self._last_pr    = None
+        self._last_out   = None
+        self._last_cov   = None
         ## Calculation thread
         self.calc_thread = None
@@ -73 +76 @@
         ## Flag to let the plug-in know that it is running standalone
         self.standalone = True
+        self._normalize_output = False
+        self._scale_output_unity = False
         
         # Log startup
@@ -265 +270 @@
             self._pr_npts= dialog.get_content()
             dialog.Destroy()
-            self.show_pr(self.pr.out, self.pr, self.pr.cov)
+            self.show_pr(self._last_out, self._last_pr, self._last_cov)
         else:
             dialog.Destroy()
@@ -285 +290 @@
 
         sum = 0.0
+        pmax = 0.0
         cov2 = numpy.ascontiguousarray(cov)
         
@@ -293 +299 @@
                 (value, dy[i]) = pr.pr_err(out, cov2, x[i])
             sum += value*pr.d_max/len(x)
+            
+            # keep track of the maximum P(r) value
+            if value>pmax:
+                pmax = value
+                
             y[i] = value
-            
-        y = y/sum
-        dy = dy/sum
+                
+        if self._normalize_output==True:
+            y = y/sum
+            dy = dy/sum
+        elif self._scale_output_unity==True:
+            y = y/pmax
+            dy = dy/pmax
         
         if cov2==None:
@@ -378 +393 @@
         for item in graph.plottables:
             if item.name==PR_FIT_LABEL:
-                return [["Add P(r) data", "Load a data file and display it on this plot", self._on_add_data],
+                m_list = [["Add P(r) data", "Load a data file and display it on this plot", self._on_add_data],
                        ["Change number of P(r) points", "Change the number of points on the P(r) output", self._on_pr_npts]]
+
+                m_list.append(["Disable P(r) scaling", 
+                               "Let the output P(r) keep the scale of the data", 
+                               self._on_disable_scaling])
+                
+                if self._scale_output_unity==False:
+                    m_list.append(["Scale P_max(r) to unity", 
+                                   "Scale P(r) so that its maximum is 1", 
+                                   self._on_scale_unity])
+                    
+                if self._normalize_output==False:
+                    m_list.append(["Normalize P(r) to unity", 
+                                   "Normalize the integral of P(r) to 1", 
+                                   self._on_normalize])
+                    
+                return m_list
+                #return [["Add P(r) data", "Load a data file and display it on this plot", self._on_add_data],
+                #       ["Change number of P(r) points", "Change the number of points on the P(r) output", self._on_pr_npts]]
 
             elif item.name==graph.selected_plottable:
@@ -386 +419 @@
         return []
 
+    def _on_disable_scaling(self, evt):
+        """
+            Disable P(r) scaling
+            @param evt: Menu event
+        """
+        self._normalize_output = False
+        self._scale_output_unity = False
+        self.show_pr(self._last_out, self._last_pr, self._last_cov)
+        
+    def _on_normalize(self, evt):
+        """
+            Switch normalization ON/OFF
+            @param evt: Menu event
+        """
+        self._normalize_output = True
+        self._scale_output_unity = False
+            
+        self.show_pr(self._last_out, self._last_pr, self._last_cov)
+        
+    def _on_scale_unity(self, evt):
+        """
+            Switch normalization ON/OFF
+            @param evt: Menu event
+        """
+        self._scale_output_unity = True
+        self._normalize_output = False
+            
+        self.show_pr(self._last_out, self._last_pr, self._last_cov)
+        
     def _on_add_data(self, evt):
         """
@@ -468 +530 @@
         # Save useful info
         self.elapsed = elapsed
+        # Keep a copy of the last result
+        self._last_pr  = pr.clone()
+        self._last_out = out
+        self._last_cov = cov
+        
         # Save Pr invertor
         self.pr = pr

prview/perspectives/pr/requirements.txt

@@ -2 +2 @@
 2. Find good defaults. Automate the inversion.
    -How do we estimate d_max? Guinier fit?
-   -How do we estimate alpha? [partially done]
+   -How do we estimate alpha? [done]
    Compute the residuals once and choose a value that brings the reg term up to 
    a fraction of the residuals.
 3. When fitting a plottable on a Graph with multiple plottables, remember the one the user chose [done]
-4. What q-range do we need to make it work? What are the constraints?
-5. Create handshake so that a plug-in can set context menu items to its own panels...
-6. Return figure of merit values and suggest the values for d_max and alpha.
+4. What q-range do we need to make it work? What are the constraints? [studied by Raiza]
+5. Create handshake so that a plug-in can set context menu items to its own panels... [done]
+6. Return figure of merit values and suggest the values for d_max and alpha. [alpha, nfunc done]
 7. Show time estimate [done]
-8. Plot the reg term alone to see what it looks like
-9. Use data loader perspective/plug-in to load data/choose file
-10. Common module for file reading and file choosing
-11. Define limits of applicability
-12. Return figures of merit
+8. Plot the reg term alone to see what it looks like 
+9. Use data loader perspective/plug-in to load data/choose file [done]
+10. Common module for file reading and file choosing 
+11. Define limits of applicability 
+12. Return figures of merit [done]
 13. Errors on output constants [done]
-14. Clean up the InversionDialog
-15. Add online help for inputs and outputs (especially figures of merit)
+14. Clean up the InversionDialog [done]
+15. Add online help for inputs and outputs (especially figures of merit) [done]
 16. Add help text to the interface (description of what we are minimizing, to explain alpha and nfunc).
-17. Get rid of all output prints.
+17. Get rid of all output prints. 
 18. Clean up
 19. Replace help dialog text with a nice image. Mention how to choose alpha and what the suggested alpha means.
 20. If a figure of merit is clearly bad, put its background red
-21. Option to save data.
-22. If you click compute before loading data, an error occurs.
-23. Plot I(q) immediately when you choose a new file.
+21. Option to save data. [done]
+22. If you click compute before loading data, an error occurs. [fixed]
+23. Plot I(q) immediately when you choose a new file. [done]
 24. *** probably need to ship the perspectives separately...
-25. If no data is available, disable the Fit button.
+25. If no data is available, disable the Fit button. [Message sent instead]
+26. Remove an added curve...
+27. Add multiple additional data curves.

prview/release_notes.txt

@@ -6 +6 @@
 Package name: None
 
-1- Version 0.2.1
+1- Version 0.2.2
 
+        - Release date: 7/8/2008
+        - Added option to normalize P(r) to 1, scale P_max to 1, or leave as is.
+        - Fixed problem with changing the number of P(r) points.
+
+   Version 0.2.1
         - Release date: 7/04/2008
         - Improved user interface.