autoscale works now also correctly in logarithmic plots (Fixes #8)

src/gui/DamarisGUI.py

@@ -2959,6 +2959,7 @@ class MonitorWidgets:
             if self.__rescale:
                 x_scale = self.display_x_scaling_combobox.get_active_text( )
                 y_scale = self.display_y_scaling_combobox.get_active_text( )
+                # x-axis
                 if x_scale == "lin":
                     self.matplot_axes.set_xscale( "linear" )
                     self.matplot_axes.set_xlim(xmin, xmax)
@@ -2971,7 +2972,7 @@ class MonitorWidgets:
                     #self.display_autoscaling_checkbutton.set_active(True) #partial fix T141
                     self.matplot_axes.set_xlim(xmax, xmin)
                     #self.matplot_axes.invert_xaxis()
-
+                # y-axis
                 if  y_scale == "lin":
                     self.matplot_axes.set_yscale( "linear" )
                     self.matplot_axes.set_ylim(ymin, ymax)
@@ -2987,8 +2988,29 @@ class MonitorWidgets:
                 if xlim_min != xmin or xlim_max != xmax:
                     self.matplot_axes.set_xlim(xmin, xmax)
 
-                # Rescale if new max is much larger than old ymax, simialar rules apply to ymin
                 ylim_min, ylim_max = self.matplot_axes.get_ylim()
+            
+                # Check if y-axis is log scale
+                yscale = self.matplot_axes.get_yscale()
+            
+                if yscale == 'log' or yscale == 'symlog':
+                    # For log scales, work with log values to avoid issues with
+                    # small numbers (undetected rescale)
+                    # Use log-space calculations
+                    log_ymin, log_ymax = numpy.log10(ymin), numpy.log10(ymax)
+
+                    log_diff = log_ymax - log_ymin
+
+                    # Add a small margin in log space (about 0.08 in log10 units ~20% margin)
+                    new_ymin = 10**(log_ymin - 0.08 * max(log_diff, 1))
+                    new_ymax = 10**(log_ymax + 0.08 * max(log_diff, 1))
+
+                    # Only update if values would change significantly
+                    if (ylim_max < ymax or ylim_min > ymin or
+                        ylim_max > new_ymax * 1.2 or ylim_min < new_ymin / 1.2):
+                        self.matplot_axes.set_ylim(new_ymin, new_ymax)
+                else:
+                    # Original linear scale logic
                     ydiff = ymax - ymin
                     if (ylim_max < ymax or ylim_min > ymin or
                         ylim_max > ymax + 0.2 * ydiff or ylim_min < ymin - 0.2 * ydiff):