script/Correlation/Correlation.py

@@ -0,0 +1,172 @@
+import math
+import sys, traceback          
+from mathutils import fit_polynomial, PolynomialFunction
+from plotutils import plot_line, plot_function
+from ch.psi.pshell.swing.Shell import STDOUT_COLOR
+import org.apache.commons.math3.stat.correlation.PearsonsCorrelation as PearsonsCorrelation
+
+start_task("outupdate", 0.0, 0.0)
+
+if get_exec_pars().source == CommandSource.ui:
+    #dx = "SINEG01-RLLE-REF10:SIG-PHASE-AVG"
+    #dy = "SINEG01-RLLE-REF20:SIG-PHASE-AVG"     
+    #dx = "SINEG01-RGUN-PUP10:SIG-AMPLT-AVG 4"
+    #dy = "SINEG01-RGUN-PUP20:SIG-AMPLT-AVG 4"
+
+    #dx = "SINDI01-RKLY-DCP10:REF-AMPLT"
+    #dy = "SINDI01-RKLY-DCP10:REF-PHASE"
+            
+    #dx = "SINDI01-RLLE-REF10:SIG-PHASE-AVG"
+    #dy = "SINDI01-RLLE-REF20:SIG-PHASE-AVG"
+
+    dx = "TESTIOC:TESTCALCOUT:Input"
+    dx = "TESTIOC:TESTCALCOUT:Output"
+
+    dy = "TESTIOC:TESTSINUS:SinCalc"
+ 
+    #dx = "SINEG01-DICT215:B1_CHARGE"
+    #dy = "SINEG01-DBPM314:Q1"     
+    
+    
+    #dx=gsx.getReadback()
+    #dy=gsy.getReadback()
+    interval = 0.1
+    window = 40
+    p = plot(None)[0]    
+    bs = False
+    linear_fit = True
+    quadratic_fit = True
+#print dx
+#print dy
+corr = None
+pars_lin = None 
+pars_quad = None
+
+
+for s in p.getAllSeries():
+    p.removeSeries(s)    
+
+_stream = None
+
+instances = []
+
+def _get_device(d):
+    global _stream
+    egu = None
+    if isinstance(d, basestring):    
+        name = d.strip()
+        d = None
+        try:
+            d = get_device(name) 
+            if d is None: 
+                d = eval(name)
+            #print name
+            if d is not None:
+                if not isinstance(r, Device): 
+                    d = None
+                else:
+                    try:
+                        egu = d.unit
+                    except:
+                        pass                            
+        except:
+            pass
+        if d is None:
+            offset = 0
+            if " " in name:
+                tokens = name.split(" ")
+                name = tokens[0]
+                offset = int(tokens[1])
+            if bs == True:
+                if _stream == None:
+                    _stream = Stream("corr_stream", dispatcher)
+                    instances.append(_stream)
+                d = _stream.addScalar(name, name, int(interval*100), offset)
+            else:
+                d =  Channel(name)    
+                instances.append(d)
+            try:
+                egu = caget(name+".EGU",'s')
+            except:
+                pass            
+    else:
+        try:
+            egu = d.unit
+        except:
+            pass            
+        
+    return d, egu
+
+dx, egux = _get_device(dx)
+dy, eguy = _get_device(dy)
+
+p.getAxis(p.AxisId.X).setLabel(egux)
+p.getAxis(p.AxisId.Y).setLabel(eguy)
+
+
+try:
+    if _stream != None:
+        _stream.initialize()
+        _stream.start(True)
+    p.addSeries(LinePlotSeries("Data")) 
+    sd=p.getSeries(0)
+    sd.setLinesVisible(False)    
+    
+    sd.setPointSize(4)
+
+    if get_exec_pars().source == CommandSource.ui:
+        if globals().has_key("marker"):
+            p.removeMarker(marker)    
+        marker=None
+            
+    
+    while(True):
+        #Sample and plot data
+        if bs == True:
+            _stream.waitValueNot(_stream.take(), 10000)
+            (x,y) = _stream.take().values
+        else:
+            x=dx.read()  
+            y=dy.read()
+        sd.appendData(x, y)
+        if len(sd.x) > window:
+            #Remove First Element
+            sd.token.remove(0)    
+        ax = sd.x
+        ay = sd.y
+        if len(ax)>2:
+            x1, x2 = min(ax), max(ax)  
+            res = (x2-x1)/100     
+            if x1!=x2:
+                #Display correlation
+                corr= PearsonsCorrelation().correlation(to_array(ax,'d'), to_array(ay,'d'))        
+                s = "Correlation=" + str(round(corr,4))
+                #print s
+                if get_exec_pars().source == CommandSource.ui:
+                    if marker is not None: 
+                        p.removeMarker(marker)
+                    marker = p.addMarker(x2+res, p.AxisId.X, s,  p.getBackground())
+                    marker.setLabelPaint(STDOUT_COLOR)
+                if linear_fit:
+                    #Calculate, print and plot linear fit
+                    pars_lin = (a0,a1) = fit_polynomial(ay, ax, 1)
+                    #print "Fit lin  a1:" , a1, " a0:",a0        
+                    y1 = poly(x1, pars_lin)
+                    y2 = poly(x2, pars_lin)
+                    plot_line(p, x1, y1, x2, y2, width = 2, color = Color.BLUE, name = "Fit Linear")
+                if quadratic_fit:
+                    #Calculate, print and plot quadratic fit
+                    pars_quad = (a0,a1,a2) = fit_polynomial(ay, ax, 2)
+                    #print "Fit quad a2:" , a2, "a1:" , a1, " a0:",a0        
+                    fitted_quad_function = PolynomialFunction(pars_quad)
+                    ax = frange(x1, x2, res, True)
+                    plot_function(p, fitted_quad_function, "Fit Quadratic", ax, color=Color.GREEN)
+        if bs != True:
+            time.sleep(interval)         
+finally:
+    for dev in instances:
+        dev.close()
+        stop_task("outupdate")
+
+
+        

script/Correlation/Correlation2.py

@@ -0,0 +1,197 @@
+import math
+import sys, traceback          
+from mathutils import fit_polynomial, PolynomialFunction
+from plotutils import plot_line, plot_function
+from ch.psi.pshell.swing.Shell import STDOUT_COLOR
+import org.apache.commons.math3.stat.correlation.PearsonsCorrelation as PearsonsCorrelation
+import ch.psi.pshell.bs.PipelineServer as PipelineServer
+
+start_task("outupdate", 0.0, 0.0)
+TYPE_CHANNEL = 0
+TYPE_STREAM = 1
+TYPE_CAMERA= 2
+
+if get_exec_pars().source == CommandSource.ui:
+    #dx = "SINDI01-RLLE-STA:SLAVE1-CPUTIMER"
+    dx = "Int16Scalar" 
+    ##dx = "SLG-LCAM-C042 x_rms"
+    dxtype = TYPE_STREAM
+    #dxtype = TYPE_CHANNEL
+    #dxtype = TYPE_CAMERA
+    #dy = "SINDI01-RLLE-STA:SLAVE1-DLTIMER"
+    #dy = "SLG-LCAM-C042 y_rms"
+    dy = "Int8Scalar"
+    dytype = TYPE_STREAM
+    #dytype = TYPE_CHANNEL
+    #dytype = TYPE_CAMERA
+    interval = 0.10
+    window = 40
+    p = plot(None)[0]    
+    linear_fit = True
+    quadratic_fit = True
+print dx, dxtype
+print dy, dytype
+
+corr = None
+pars_lin = None 
+pars_quad = None
+
+
+
+bs = TYPE_STREAM in [dxtype, dytype]
+
+
+for s in p.getAllSeries():
+    p.removeSeries(s)    
+
+_stream = None
+_camname = None
+
+instances = []
+
+def _get_device(d, type):
+    global _stream, _camname
+    egu = None
+    if isinstance(d, basestring):    
+        name = d.strip()
+        d = None
+        try:
+            d = get_device(name) 
+            if d is None: 
+                d = eval(name)
+            #print name
+            if d is not None:
+                if not isinstance(r, Device): 
+                    d = None
+                else:
+                    try:
+                        egu = d.unit
+                    except:
+                        pass                            
+        except:
+            pass
+        if d is None:            
+            offset = 0
+            
+            if type==TYPE_STREAM:
+                if " " in name:
+                    tokens = name.split(" ")
+                    name = tokens[0]
+                    offset = int(tokens[1])
+                if _stream == None:
+                    _stream = Stream("corr_stream", dispatcher)
+                    instances.append(_stream)
+                d = _stream.addScalar(name, name, int(interval*100), offset)
+            elif type==TYPE_CHANNEL:
+                d =  Channel(name)    
+                d.set_monitored(True)
+            elif  type==TYPE_CAMERA:
+                tokens = name.split(" ")
+                _camname = tokens[0]
+                field = tokens[1]
+                return field, ""
+            else:
+                raise Exception("Invalid type: " + str(type))
+            if not isinstance(d, basestring):    
+                instances.append(d)
+            try:
+                egu = caget(name+".EGU",'s')
+            except:
+                pass            
+    else:
+        try:
+            egu = d.unit
+        except:
+            pass                    
+    return d, egu
+
+dx, egux = _get_device(dx, dxtype)
+dy, eguy = _get_device(dy, dytype)
+
+p.getAxis(p.AxisId.X).setLabel(egux)
+p.getAxis(p.AxisId.Y).setLabel(eguy)
+
+
+try:
+    if _stream != None:
+        _stream.initialize()
+        _stream.start(True)
+    if _camname != None:
+        cam_server.start(_camname )
+        cam_server.stream.waitCacheChange(10000);
+        if  dxtype==TYPE_CAMERA:
+            dx=cam_server.stream.getChild(dx)
+        if  dytype==TYPE_CAMERA:
+            dy=cam_server.stream.getChild(dy)
+        
+    p.addSeries(LinePlotSeries("Data")) 
+    sd=p.getSeries(0)
+    sd.setLinesVisible(False)    
+    
+    sd.setPointSize(4)
+    
+    if get_exec_pars().source == CommandSource.ui:
+        if globals().has_key("marker"):
+            p.removeMarker(marker)    
+        marker=None
+            
+    
+    while(True):
+        #Sample and plot data
+        if bs == True:
+            _stream.waitValueNot(_stream.take(), 10000)
+            bsdata = list(_stream.take().values)
+        
+        if dxtype==TYPE_CHANNEL:
+            x=dx.read() 
+        elif dxtype==TYPE_STREAM:
+            x=bsdata.pop(0)
+        elif dxtype==TYPE_CAMERA:
+            x=dx.read()
+                        
+        if dytype==TYPE_CHANNEL:
+            y=dy.read()
+        elif dytype==TYPE_STREAM:
+            y=bsdata.pop(0)
+        elif dytype==TYPE_CAMERA:
+            y=dy.read()
+        
+        sd.appendData(x, y)
+        if len(sd.x) > window:
+            #Remove First Element
+            sd.token.remove(0)    
+        ax = sd.x
+        ay = sd.y
+        if len(ax)>2:
+            x1, x2 = min(ax), max(ax)  
+            res = (x2-x1)/100     
+            if x1!=x2:
+                #Display correlation
+                corr= PearsonsCorrelation().correlation(to_array(ax,'d'), to_array(ay,'d'))        
+                s = "Correlation=" + str(round(corr,4))
+                #print s
+                if get_exec_pars().source == CommandSource.ui:
+                    if marker is not None: 
+                        p.removeMarker(marker)
+                    marker = p.addMarker(x2+res, p.AxisId.X, s,  p.getBackground())
+                    marker.setLabelPaint(STDOUT_COLOR)
+                if linear_fit:
+                    #Calculate, print and plot linear fit
+                    pars_lin = (a0,a1) = fit_polynomial(ay, ax, 1)
+                    #print "Fit lin  a1:" , a1, " a0:",a0        
+                    y1 = poly(x1, pars_lin)
+                    y2 = poly(x2, pars_lin)
+                    plot_line(p, x1, y1, x2, y2, width = 2, color = Color.BLUE, name = "Fit Linear")
+                if quadratic_fit:
+                    #Calculate, print and plot quadratic fit
+                    pars_quad = (a0,a1,a2) = fit_polynomial(ay, ax, 2)
+                    #print "Fit quad a2:" , a2, "a1:" , a1, " a0:",a0        
+                    fitted_quad_function = PolynomialFunction(pars_quad)
+                    ax = frange(x1, x2, res, True)
+                    plot_function(p, fitted_quad_function, "Fit Quadratic", ax, color=Color.GREEN)
+        if bs != True:
+            time.sleep(interval)         
+finally:
+    for dev in instances:
+        dev.close()
+    stop_task("outupdate")