Fixed issue with q binning. When config file has been saved with line…

…ar binning, new binning stay linear, even if application does not say so.. this has been fixed and will always be linear for now on. This refs #4303

Code/Mantid/Framework/PythonAPI/PythonAlgorithms/RefLReduction.py

@@ -124,10 +124,16 @@ def PyExec(self):
         # Steps for TOF rebin
         TOFsteps = 100.0
 
+        #use now a global q binning (user does not have control over it)
+        #q_min = 0.005
+        #q_step = -0.01
+
         # Q binning for output distribution
         q_min = self.getProperty("QMin")
         q_step = self.getProperty("QStep")
-
+        if (q_step > 0):
+            q_step = -q_step
+
         #dimension of the detector (256 by 304 pixels)
         maxX = 304
         maxY = 256

Code/Mantid/scripts/Interface/reduction_gui/widgets/reflectometer/base_ref_reduction.py

@@ -284,6 +284,7 @@ def _smooth_x_axis(self, x_axis, y_axis, e_axis):
                     _y = 0.
                     _e = 0.
                 else:
+
                     _error = 1./_left_e2 + 1./_right_e2
                     _x = (_left_x + _right_x) / 2. 
                     _y = (_left_y/_left_e2 + _right_y/_right_e2) / _error 
@@ -338,6 +339,19 @@ def _create_ascii_clicked(self):
             text = [line1, line2, line3]
         else:
             text = ['#Q(1/Angstrom) R delta_R']
+
+#        #rebinned using output factors
+#        q_min = float(self._summary.q_min_edit.text())
+#        q_bin = -float(self._summary.q_step_edit.text())
+#
+#        mt = mtd['ref_combined']
+#        x_axis = mt.readX(0)[:]
+#        q_max = float(x_axis[-1]) 
+#        
+#        q_binning = [q_min, q_bin, q_max]
+#        Rebin(InputWorkspace='ref_combined',
+#              OutputWorkspace='ref_combined',
+#              Params=q_binning)
 
         mt = mtd['ref_combined']
         x_axis = mt.readX(0)[:]