def testIntegers(self):
print 'test integers get and set'
a = np.array([8, 9, 10, 11, 12, 13])
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a.shape = 3, 2
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a[np.where(a > 10)] = -1
self.checkitems([[8, 9], [10, -1], [-1, -1]], a)
# fancy indexing
a = np.array([8, 9, 10, 11, 12, 13])
a.shape = 3, 2
self.assertEquals(13, a[(2, 1)])
d = a[(2, 1), ]
self.checkitems([[12, 13], [10, 11]], d)
tm = np.array(self.mm)
d = tm[[1]]
self.checkitems([[[20., 30.], [42., 56.]]], d)
d = tm[np.array([0, 1]), np.array([1, 1])]
self.checkitems([[6., 12.], [42., 56.]], d)
d = tm[np.array([0, 1]), np.array([1, 1]), :1]
self.checkitems([[6.], [42.]], d)
d = tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
d = tm[[[0, 1], [1, 1]], [[1, 1], [1, 0]], :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1] = -2.3
self.checkitems([[[0., 2.], [-2.3, 12.]], [[-2.3, 30.], [-2.3, 56.]]],
tm)
# fancy with booleans too
tm = np.array(self.mm)
d = tm[np.array([0, 1]), np.array([False, True]), :1]
self.checkitems([[6.], [42.]], d)
# TODO broadcasting with lower ranked booleam
d = np.arange(8).reshape(4, 2)
m = np.array([False, True, True, False])
print 'broadcast', d[m]
self.checkitems([[2, 3], [4, 5]], d[m])
def testIntegers(self):
print('test integers get and set')
a = np.array([8,9,10,11,12,13])
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a.shape = 3,2
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a[np.where(a > 10)] = -1
self.checkitems([[8,9],[10,-1], [-1, -1]], a)
# fancy indexing
a = np.array([8,9,10,11,12,13])
a.shape = 3,2
self.assertEquals(13, a[(2,1)])
d = a[(2,1),]
self.checkitems([[12, 13], [10, 11]], d)
tm = np.array(self.mm)
d = tm[[1]]
self.checkitems([[[20., 30.], [42., 56.]]], d)
d = tm[np.array([0, 1]), np.array([1, 1])]
self.checkitems([[6., 12.], [42., 56.]], d)
d = tm[np.array([0, 1]), np.array([1, 1]), :1]
self.checkitems([[6.], [42.]], d)
d = tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
d = tm[[[0, 1], [1, 1]], [[1, 1], [1, 0]], :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1] = -2.3
self.checkitems([[[0., 2.], [-2.3, 12.]], [[-2.3, 30.], [-2.3, 56.]]], tm)
# fancy with booleans too
tm = np.array(self.mm)
d = tm[np.array([0, 1]), np.array([False, True]), :1]
self.checkitems([[6.], [42.]], d)
# TODO broadcasting with lower ranked booleam
d = np.arange(8).reshape(4,2)
m = np.array([False, True, True, False])
print('broadcast', d[m])
self.checkitems([[2, 3], [4, 5]], d[m])
def find_peak_coords(peak_image):
data = peak_image[0].transpose()
roi = data[centre[0]-width/2:centre[0]+width/2][:]
peak_value = np.max(roi)
coords = dnp.where(peak_image[0]==peak_value)
coords = zip(coords[0], coords[1])
return coords[0]
def testIntegers(self):
print 'test integers get and set'
a = np.array([8,9,10,11,12,13])
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a.shape = 3,2
d = a[np.where(a > 10)]
self.checkitems([11, 12, 13], d)
a[np.where(a > 10)] = -1
self.checkitems([[8,9],[10,-1], [-1, -1]], a)
# fancy indexing
a = np.array([8,9,10,11,12,13])
a.shape = 3,2
self.assertEquals(13, a[(2,1)])
d = a[(2,1),]
self.checkitems([[12, 13], [10, 11]], d)
tm = np.array(self.mm)
d = tm[[1]]
self.checkitems([[[20., 30.], [42., 56.]]], d)
d = tm[np.array([0, 1]), np.array([1, 1])]
self.checkitems([[6., 12.], [42., 56.]], d)
d = tm[np.array([0, 1]), np.array([1, 1]), :1]
self.checkitems([[6.], [42.]], d)
d = tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
d = tm[[[0, 1], [1, 1]], [[1, 1], [1, 0]], :1]
self.checkitems([[[6.], [42.]], [[42.], [20.]]], d)
tm[np.array([[0, 1], [1, 1]]), np.array([[1, 1], [1, 0]]), :1] = -2.3
self.checkitems([[[0., 2.], [-2.3, 12.]], [[-2.3, 30.], [-2.3, 56.]]], tm)
# fancy with booleans too
tm = np.array(self.mm)
d = tm[np.array([0, 1]), np.array([False, True]), :1]
self.checkitems([[6.], [42.]], d)
def onUpdatePlot(self, event=None):
if self.paused or not self.needs_refresh:
return
tnow = time.time()
# set timescale sec/min/hour
timescale = 1.0
if self.time_choice.GetSelection() == 1:
timescale = 1. / 60
elif self.time_choice.GetSelection() == 2:
timescale = 1. / 3600
ylabelset, y2labelset = False, False
xlabel = 'Elapsed Time (%s)' % self.timelabel
itrace = -1
update_failed = False
hasplot = False
span1 = (1, 0)
did_update = False
left_axes = self.plotpanel.axes
right_axes = self.plotpanel.get_right_axes()
for irow, pname, uselog, color, ymin, ymax in self.get_current_traces(
):
if pname not in self.pvdata:
continue
itrace += 1
if len(self.plots_drawn) < itrace:
self.plots_drawn.extend([False] * 3)
side = 'left'
if itrace == 1:
side = 'right'
data = self.pvdata[pname][:]
if len(data) < 2:
update_failed = True
continue
tdat = timescale * (array([i[0] for i in data]) - tnow)
mask = where(tdat > self.tmin)
if (len(mask[0]) < 2
or ((abs(min(tdat)) / abs(1 - self.tmin)) > 0.1)):
data.append((time.time(), data[0][-1]))
tdat = timescale * (array([i[0] for i in data]) - tnow)
mask = where(tdat > self.tmin)
i0 = mask[0][0]
if i0 > 0:
i0 = i0 - 1
i1 = mask[0][-1] + 1
tdat = timescale * (array([i[0] for i in data[i0:i1]]) - tnow)
ydat = array([i[1] for i in data[i0:i1]])
if len(ydat) < 2:
update_failed = True
continue
if ymin is None:
ymin = min(ydat)
if ymax is None:
ymax = max(ydat)
# for more that 2 plots, scale to left hand axis
if itrace == 0:
span1 = (ymax - ymin, ymin)
if span1[0] * ymax < 1.e-6:
update_failed = True
continue
elif itrace > 1:
yr = abs(ymax - ymin)
if yr > 1.e-9:
ydat = span1[1] + 0.99 * (ydat - ymin) * span1[0] / yr
ymin, ymax = min(ydat), max(ydat)
if self.needs_refresh:
if itrace == 0:
self.plotpanel.set_ylabel(pname)
elif itrace == 1:
self.plotpanel.set_y2label(pname)
if not self.plots_drawn[itrace]:
plot = self.plotpanel.oplot
if itrace == 0:
plot = self.plotpanel.plot
try:
plot(tdat,
ydat,
drawstyle='steps-post',
side=side,
ylog_scale=uselog,
color=color,
xmin=self.tmin,
xmax=0,
xlabel=xlabel,
label=pname,
autoscale=False)
self.plots_drawn[itrace] = True
except:
update_failed = True
else:
try:
self.plotpanel.update_line(itrace,
tdat,
ydat,
draw=False)
self.plotpanel.set_xylims((self.tmin, 0, ymin, ymax),
side=side,
autoscale=False)
did_update = True
except:
update_failed = True
axes = left_axes
if itrace == 1:
axes = right_axes
if uselog and min(ydat) > 0:
axes.set_yscale('log', basey=10)
else:
axes.set_yscale('linear')
self.plotpanel.set_title(
time.strftime("%Y-%b-%d %H:%M:%S", time.localtime()))
if did_update:
self.plotpanel.canvas.draw()
self.needs_refresh = update_failed
return
def refl(runfiles, pathtofiles, outputpath, scalar, beamheight, footprint, angularfudgefactor, wl, back=()):
# scalar - scale factor to divide the data by
# beamheight FWHM in microns
# footprint in mm
# angular offset correction in degrees
# wavelength in wl
# back is an optional variable to subtract a background, set back=1 to do a background subtraction
qq = []
RR = []
dR = []
ii = -1
for filename in runfiles:
data = dnp.io.load(pathtofiles + "/" + str(filename) + ".dat")
ii += 1
theta = data.alpha
# work out the q vector
qqtemp = 4 * dnp.pi * dnp.sin((theta + angularfudgefactor) * dnp.pi /180) / wl
#qqtemp = data.qdcd
# this section is to allow users to set limits on the q range used from each file
if not 'qmin' + str(ii) in refl.__dict__:
qmin = qqtemp.min()
else:
print "USER SET",
qmin = refl.__getattribute__('qmin' + str(ii))
print 'refl.qmin' + str(ii) + " = " + str(qmin) + " ;",
if not 'qmax' + str(ii) in refl.__dict__:
qmax = qqtemp.max()
else:
print "USER SET",
qmax = refl.__getattribute__('qmax' + str(ii))
print 'refl.qmax' + str(ii) + " = " + str(qmax) + " ;",
roi1_sum = data.roi1_sum
roi1_sum = roi1_sum[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
roi1dr = dnp.sqrt(roi1_sum)
theta = theta[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
qqtemp = qqtemp[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
bg_sum = dnp.zeros(len(roi1_sum))
bg_dr = dnp.zeros(len(roi1dr))
# if background ROI number given as int, convert to a single-item tuple
if type(back) == int:
back = (back,)
# subtract any background ROIs from the data
if len(back) > 0 and back[0] > 0:
if ii==0:
print "Using background from " + str(len(back)) + " ROIs: " + str(back)
for bg in back:
if ('roi' + str(bg) + '_sum' in data.keys()):
bg_cur = data[data.keys().index('roi' +str(bg) + '_sum')]
dr_cur = dnp.sqrt(bg_cur)
(bg_sum, bg_dr) = ep.EPadd(bg_sum, bg_dr, bg_cur, dr_cur)
(bg_sum, bg_dr) = ep.EPmulk(bg_sum, bg_dr, 1.0/len(back))
else:
if ii==0:
print "Not subtracting a background"
(RRtemp, drtemp) = ep.EPsub(roi1_sum, roi1dr, bg_sum, bg_dr)
# do a footprint correction.
# assumes that the beam is gaussian in profile, with a FWHM of "beamheight".
# footprint of sample is measured in mm.
areamultiplier = 2*(norm.cdf(footprint * dnp.sin((theta + angularfudgefactor) / 180 * dnp.pi) / 2, 0, 1e-3 * beamheight/ (2*dnp.sqrt(2*dnp.log(2)))) - 0.5)
RRtemp /= areamultiplier
drtemp /= areamultiplier
# for the 2nd, 3rd, 4th q ranges have to splice the data on the end of the preexisting data
if(ii > 0):
# splice
(scalingfactor, sferror) = nsplice.getScalingInOverlap(qq, RR, dR, qqtemp, RRtemp, drtemp)
RRtemp *= scalingfactor
drtemp *= scalingfactor
print "Error in scaling factor: %2.3f %%" % (sferror/scalingfactor*100)
# now concatenate the data.
qq = dnp.concatenate((qq, qqtemp))
RR = dnp.concatenate((RR, RRtemp))
dR = dnp.concatenate((dR, drtemp))
# end of per-file loop
RR /= scalar
dR /= scalar
RR = RR[np.argsort(qq)]
dR = dR[np.argsort(qq)]
qq = np.sort(qq)
# write out the data.
np.savetxt(outputpath+"/"+str(runfiles[0])+"_refl.dat",dnp.concatenate((qq,RR,dR)).reshape(3,qq.shape[0]).transpose())
def testWhere(self):
print 'test where'
t = np.array([[[-0.5, 0.1], [2, 0]], [[1.3, 1.4], [-10, -20]]])
tm = np.where(t > 0, np.array(self.mm), -2.3)
self.checkitems([ [[-2.3, 2.], [6., -2.3]], [[20., 30.], [-2.3, -2.3]] ], tm)
def testWhere(self):
print 'test where'
t = np.array([[[-0.5, 0.1], [2, 0]], [[1.3, 1.4], [-10, -20]]])
tm = np.where(t > 0, np.array(self.mm), -2.3)
self.checkitems([[[-2.3, 2.], [6., -2.3]], [[20., 30.], [-2.3, -2.3]]],
tm)
def refl(runfiles,
pathtofiles,
outputpath,
scalar,
beamheight,
footprint,
angularfudgefactor,
wl,
back=()):
# scalar - scale factor to divide the data by
# beamheight FWHM in microns
# footprint in mm
# angular offset correction in degrees
# wavelength in wl
# back is an optional variable to subtract a background, set back=1 to do a background subtraction
qq = []
RR = []
dR = []
ii = -1
for filename in runfiles:
data = dnp.io.load(pathtofiles + "/" + str(filename) + ".dat")
ii += 1
theta = data.alpha
# work out the q vector
qqtemp = 4 * dnp.pi * dnp.sin(
(theta + angularfudgefactor) * dnp.pi / 180) / wl
#qqtemp = data.qdcd
# this section is to allow users to set limits on the q range used from each file
if not 'qmin' + str(ii) in refl.__dict__:
qmin = qqtemp.min()
else:
print "USER SET",
qmin = refl.__getattribute__('qmin' + str(ii))
print 'refl.qmin' + str(ii) + " = " + str(qmin) + " ;",
if not 'qmax' + str(ii) in refl.__dict__:
qmax = qqtemp.max()
else:
print "USER SET",
qmax = refl.__getattribute__('qmax' + str(ii))
print 'refl.qmax' + str(ii) + " = " + str(qmax) + " ;",
roi1_sum = data.roi1_sum
roi1_sum = roi1_sum[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
roi1dr = dnp.sqrt(roi1_sum)
theta = theta[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
qqtemp = qqtemp[dnp.where((qqtemp >= qmin) & (qqtemp <= qmax))]
bg_sum = dnp.zeros(len(roi1_sum))
bg_dr = dnp.zeros(len(roi1dr))
# if background ROI number given as int, convert to a single-item tuple
if type(back) == int:
back = (back, )
# subtract any background ROIs from the data
if len(back) > 0 and back[0] > 0:
if ii == 0:
print "Using background from " + str(
len(back)) + " ROIs: " + str(back)
for bg in back:
if ('roi' + str(bg) + '_sum' in data.keys()):
bg_cur = data[data.keys().index('roi' + str(bg) + '_sum')]
dr_cur = dnp.sqrt(bg_cur)
(bg_sum, bg_dr) = ep.EPadd(bg_sum, bg_dr, bg_cur, dr_cur)
(bg_sum, bg_dr) = ep.EPmulk(bg_sum, bg_dr, 1.0 / len(back))
else:
if ii == 0:
print "Not subtracting a background"
(RRtemp, drtemp) = ep.EPsub(roi1_sum, roi1dr, bg_sum, bg_dr)
# do a footprint correction.
# assumes that the beam is gaussian in profile, with a FWHM of "beamheight".
# footprint of sample is measured in mm.
areamultiplier = 2 * (norm.cdf(
footprint * dnp.sin(
(theta + angularfudgefactor) / 180 * dnp.pi) / 2, 0,
1e-3 * beamheight / (2 * dnp.sqrt(2 * dnp.log(2)))) - 0.5)
RRtemp /= areamultiplier
drtemp /= areamultiplier
# for the 2nd, 3rd, 4th q ranges have to splice the data on the end of the preexisting data
if (ii > 0):
# splice
(scalingfactor,
sferror) = nsplice.getScalingInOverlap(qq, RR, dR, qqtemp, RRtemp,
drtemp)
RRtemp *= scalingfactor
drtemp *= scalingfactor
print "Error in scaling factor: %2.3f %%" % (sferror /
scalingfactor * 100)
# now concatenate the data.
qq = dnp.concatenate((qq, qqtemp))
RR = dnp.concatenate((RR, RRtemp))
dR = dnp.concatenate((dR, drtemp))
# end of per-file loop
RR /= scalar
dR /= scalar
RR = RR[np.argsort(qq)]
dR = dR[np.argsort(qq)]
qq = np.sort(qq)
# write out the data.
np.savetxt(
outputpath + "/" + str(runfiles[0]) + "_refl.dat",
dnp.concatenate((qq, RR, dR)).reshape(3, qq.shape[0]).transpose())
