various improvements (i.e. i forgot ...)

alignment.py

@@ -16,6 +16,7 @@ import time
 
 #defaultE = np.arange(1024)*0.12+7060
 defaultE = (np.arange(1024)-512)*0.189+7123
+defaultE = 7063.5+0.1295*np.arange(1024); # done on nov 30 2016, based on xppl3716:r77
 
 __i = np.arange(1024)
 __x = (__i-512)/512
@@ -83,7 +84,7 @@ fit_ret = collections.namedtuple("fit_ret",["init_pars","final_pars",\
 def calcFOM(p1,p2,ratio):
   idx = ( p1>p1.max()/10 )# & (p2>p2.max()/10)
   ratio = ratio[idx]
-  return ratio.std()/ratio.mean()
+  return ratio.std()/np.abs(ratio.mean())
   
 def subtractBkg(imgs,nPix=100,bkg_type="line"):
   if imgs.ndim == 2: imgs = imgs[np.newaxis,:]
@@ -265,6 +266,26 @@ def saveAlignment(fname,transform,roi1,roi2):
 def loadAlignment(fname):
   return np.load(fname).item()
 
+def getBestTransform(results):
+  # try to see if it is unravelled
+  if not hasattr(results,"_fields"): results = unravel_results(results)
+  # find best based on FOM
+  idx = np.nanargmin(np.abs(results.fom))
+  parnames = results.init_pars.keys()
+  bestPars = dict()
+  for n in parnames:
+    bestPars[n] = results.final_pars[n][idx]
+    if n.find("limit_") == 0:
+      if bestPars[n][0] is None:
+        bestPars[n] = None
+      else:
+        bestPars[n] = tuple(bestPars[n])
+  return bestPars,np.nanmin(np.abs(results.fom))
+  
+  
+  
+  
+
 def unravel_results(res,nSaveImg='all'):
   final_pars = dict()
   # res[0].fit_result is a list if we are trying to unravel retult from getShots
@@ -275,8 +296,12 @@ def unravel_results(res,nSaveImg='all'):
   final_pars = dict()
   init_pars  = dict()
   for n in parnames:
+    if n.find("limit_") == 0:
+      final_pars[n] = np.vstack( [r.final_pars[n] for r in res])
+      init_pars[n]  = np.vstack( [r.init_pars[n] for r in res] )
+    else:
       final_pars[n] = np.hstack( [r.final_pars[n] for r in res])
-    init_pars[n] = np.hstack( [r.init_pars[n] for r in res])
+      init_pars[n]  = np.hstack( [r.init_pars[n] for r in res] )
   im1  = np.asarray( [r.im1 for r in res if r.im1.shape[0] != 0])
   im2  = np.asarray( [r.im2 for r in res if r.im2.shape[0] != 0])
   if nSaveImg != 'all':
@@ -582,9 +607,9 @@ def doShots(imgs1,imgs2,initpars,nJobs=16,doFit=False,returnBestTransform=False,
     (joblib.delayed(doShot)(imgs1[i],imgs2[i],initpars,doFit=doFit) for i in range(N))
   res = unravel_results(pool,nSaveImg=nSaveImg)
   if returnBestTransform:
-    idx = np.nanargmin( res.fom );
-    print("FOM for best alignment %.2f"%res.fom[idx])
-    return res,pool[idx].final_pars
+    bestT,fom = getBestTransform(res)
+    print("FOM for best alignment %.2f"%fom)
+    return res,bestT
   else:
     return res
 #  out = collections.OrderedDict( enumerate(pool) )

xanes_analyzeRun.py

@@ -221,6 +221,50 @@ class AnalyzeRun(object):
     self.initAlign = gui.transform
     gui.save(fname)
 
+  def doShot(self,shot=0,calib=None,initpars=None,im1=None,im2=None,doFit=True,show=False,showInit=False,save=False,savePlot="auto"):
+    if initpars is None: initpars= self.initAlign
+    if (im1 is None) or (im2 is None):
+      im1,im2 = self.getShots(shot,calib=calib); im1=im1[0]; im2=im2[0]
+    r = alignment.doShot(im1,im2,initpars,doFit=doFit,show=showInit)
+    im1 = r.im1
+    im2 = r.im2
+    self.initAlign = r.final_pars
+    if show:
+      if savePlot == "auto":
+        if not os.path.isdir(g_folder_out): os.makedirs(g_folder_out)
+        savePlot = g_folder_out+"/run%04d_calib%s_shot%04d_fit.png" % (self.run,calib,shot)
+      alignment.plotShot(im1,im2,res=r,save=savePlot)
+    if save: self.saveTransform()
+    return r
+
+  def doShots(self,shots=slice(0,50),calib=None,initpars=None,doFit=False,returnBestTransform=False,nSaveImg=5,nInChunks=250):
+    """
+        shots   : slice to define shots to read, use 'all' for all shots in calibcycle
+        nSaveImg : save saveImg images in memory (self.results), use 'all' for all
+                                 useful for decreasing memory footprint
+    """
+    if initpars is None: initpars= self.initAlign
+    if shots == "all": shots = list(range(self.nShotsPerCalib[calib]))
+    if isinstance(shots,slice):
+      nmax = self.nShotsPerCalib[calib] if calib is not None else self.data.spec1.nShots
+      shots = sliceToIndices(shots,nmax)
+    chunks = x3py.toolsVarious.chunk(shots,nInChunks)
+    ret_chunks = []
+    for chunk in chunks:
+      s1,s2 = self.getShots(chunk,calib=calib)
+      ret_chunk = alignment.doShots(s1,s2,initpars=initpars,doFit=doFit,\
+                   returnBestTransform=False,nSaveImg=nSaveImg)
+      ret_chunks.append(ret_chunk)
+      if nSaveImg != 'all' and len(chunk)<nSaveImg: nSaveImg = 0
+    ret = alignment.unravel_results(ret_chunks)
+    bestT,fom = alignment.getBestTransform(ret)
+    if doFit: self.initAlign = bestT
+    # keep it for later !
+    self.results[calib] = ret
+    if returnBestTransform:
+      return ret,bestT
+    else:
+      return ret
 
   def analyzeScan(self,initpars=None,nShotsPerCalib="all",calibs="all",calibsToFit="all",nImagesToFit=0,nSaveImg=5):
     """ nImagesToFit: number of images to Fit per calibcycle, (int or "all")
@@ -246,7 +290,8 @@ class AnalyzeRun(object):
                 initpars=initpars,nSaveImg=nSaveImg,returnBestTransform=True);
         initpars = bestTransf; self.initAlign=bestTransf
       if nToFit < len(shots):
-        ret2 = self.doShots(shots[nToFit:],initpars=initpars,doFit=False,nSaveImg=0)
+        ret2 = self.doShots(shots[nToFit:],calib=calib,initpars=initpars,
+                doFit=False,nSaveImg=0)
         if ret is None:
           ret = ret2
         else:
@@ -257,50 +302,6 @@ class AnalyzeRun(object):
       print("Calib cycle %d/%d -> %.3f (best FOM: %.2f)" % (ic,nC,self.scanpos[ic],np.nanmin(ret.fom)))
     return [self.results[c] for c in calibs]
 
-  def doShot(self,shot=0,calib=None,initpars=None,im1=None,im2=None,doFit=True,show=False,showInit=False,save=False,savePlot="auto"):
-    if initpars is None: initpars= self.initAlign
-    if (im1 is None) or (im2 is None):
-      im1,im2 = self.getShots(shot,calib=calib); im1=im1[0]; im2=im2[0]
-    r = alignment.doShot(im1,im2,initpars,doFit=doFit,show=showInit)
-    im1 = r.im1
-    im2 = r.im2
-    self.initAlign = r.final_pars
-    if show:
-      if savePlot == "auto":
-        if not os.path.isdir(g_folder_out): os.makedirs(g_folder_out)
-        savePlot = g_folder_out+"/run%04d_calib%s_shot%04d_fit.png" % (self.run,calib,shot)
-      alignment.plotShot(im1,im2,res=r,save=savePlot)
-    if save: self.saveTransform()
-    return r
-
-  def doShots(self,shots=slice(0,50),calib=None,initpars=None,doFit=False,returnBestTransform=False,nSaveImg='all',nInChunks=250):
-    """
-        shots   : slice to define shots to read, use 'all' for all shots in calibcycle
-        nSaveImg : save saveImg images in memory (self.results), use 'all' for all
-                                 useful for decreasing memory footprint
-    """
-    if initpars is None: initpars= self.initAlign
-    if shots == "all": shots = list(range(slice(self.nShotsPerCalib[calib])))
-    if isinstance(shots,slice):
-      nmax = self.nShotsPerCalib[calib] if calib is not None else self.data.spec1.nShots
-      shots = sliceToIndices(shots,nmax)
-    chunks = x3py.toolsVarious.chunk(shots,nInChunks)
-    ret = []
-    for chunk in chunks:
-      s1,s2 = self.getShots(chunk,calib=calib)
-      temp = alignment.doShots(s1,s2,initpars=initpars,doFit=doFit,\
-            returnBestTransform=False,nSaveImg=nSaveImg)
-      ret.append(temp)
-    ret = alignment.unravel_results(ret)
-    idxBest = np.nanargmin(ret.fom)
-    transformForBestFit = dict( [ (p,ret.final_pars[p][idxBest]) for p in ret.final_pars.keys() ] )
-    if doFit: self.initAlign = transformForBestFit
-    # keep it for later !
-    self.results[calib] = ret
-    if returnBestTransform:
-      return ret,transformForBestFit
-    else:
-      return ret
 
   def save(self,fname="auto",overwrite=False):
     if len(self.results) == 0: