def morph(face1, face1pts, face2, face2pts, warp_frac=0.5, dissolve_frac=0.5):
#we assume face1pts and face2pts contains the corners of the image
face1 = np.copy(face1)
face2 = np.copy(face2)
avgpts = (1-warp_frac)*face1pts + warp_frac*face2pts
avgpts[:4] = np.trunc(avgpts[:4])
face = np.zeros((avgpts[3,0], avgpts[3,1], 3))
delaunay_triangulation = Delaunay(avgpts)
simplices = delaunay_triangulation.simplices
triang1 = [[face1pts[s[0]], face1pts[s[1]], face1pts[s[2]]] for s in simplices]
triang2 = [[face2pts[s[0]], face2pts[s[1]], face2pts[s[2]]] for s in simplices]
triang = [[avgpts[s[0]], avgpts[s[1]], avgpts[s[2]]] for s in simplices]
affine_t1 = [compute_transform(triang[i], triang1[i]) for i in range(len(triang1))]
affine_t2 = [compute_transform(triang[i], triang2[i]) for i in range(len(triang2))]
for y in range(face.shape[0]):
for x in range(face.shape[1]):
trinum1 = tsearch((y, x), triang1)
vec1 = np.dot(affine_t1[trinum1], np.array([x, y, 1]))
vec1 = np.trunc(vec1)
trinum2 = tsearch((y, x), triang2)
vec2 = np.dot(affine_t2[trinum2], np.array([x, y, 1]))
vec2 = np.trunc(vec2)
try:
face[y,x,:] = (1-dissolve_frac)*face1[vec1[1]-1,vec1[0]-1] + dissolve_frac*face2[vec2[1]-1,vec2[0]-1]
except:
print (vec1[1]-1,vec1[0]-1), (vec2[1]-1,vec2[0]-1)
return face
def set_jds(self, val1, val2):
self._check_scale(self._scale) # Validate scale.
sum12, err12 = two_sum(val1, val2)
iy_start = np.trunc(sum12).astype(np.int)
extra, y_frac = two_sum(sum12, -iy_start)
y_frac += extra + err12
val = (val1 + val2).astype(np.double)
iy_start = np.trunc(val).astype(np.int)
imon = np.ones_like(iy_start)
iday = np.ones_like(iy_start)
ihr = np.zeros_like(iy_start)
imin = np.zeros_like(iy_start)
isec = np.zeros_like(y_frac)
# Possible enhancement: use np.unique to only compute start, stop
# for unique values of iy_start.
scale = self.scale.upper().encode('ascii')
jd1_start, jd2_start = erfa.dtf2d(scale, iy_start, imon, iday,
ihr, imin, isec)
jd1_end, jd2_end = erfa.dtf2d(scale, iy_start + 1, imon, iday,
ihr, imin, isec)
t_start = Time(jd1_start, jd2_start, scale=self.scale, format='jd')
t_end = Time(jd1_end, jd2_end, scale=self.scale, format='jd')
t_frac = t_start + (t_end - t_start) * y_frac
self.jd1, self.jd2 = day_frac(t_frac.jd1, t_frac.jd2)
def init_boid(amaze):
pos = np.zeros(cfg.Dimensions, dtype=np.float32)
# Start from a random cell
pos[0] = np.float32(np.trunc(np.random.uniform(0, cfg.maze_width)))
pos[1] = np.float32(np.trunc(np.random.uniform(0, cfg.maze_height)))
# Ensure that we are not placing the boid into a wall ---------------------
# Change position until hit a free cell
while isinstance(
amaze.matrix[int(pos[0])][int(pos[1])].shape,
maze.Wall):
# While the cell is filled with a wall (==1)
pos[0] = np.float32(np.trunc(np.random.uniform(0, cfg.maze_width)))
pos[1] = np.float32(np.trunc(np.random.uniform(0, cfg.maze_height)))
# Check that we are not placing the boind into the wall
# Move boid to the center of the cell
pos[0] += 0.5
pos[1] += 0.5
vel = np.zeros(cfg.Dimensions, dtype=np.float32)
# TODO: change pheromone_level type to uint16
pheromone_level = np.zeros([1], dtype=np.float32)
return np.concatenate((pos, vel, pheromone_level))
def split_data(dataset):
'''
Split the dataset into training, validation, and test set. The split will be 70, 20, 10
:param dataset: the entire dataset containing positive and negative images. each row is a tuple containing the data and label
:param pixel_res: the number of neurons per image
:return: a tuple of training, validation, and test
'''
# shuffle the data to randomize
shuffled = np.random.permutation(dataset)
# get the sizes based on split percentage for each set
print 'total dataset size:', len(shuffled)
ts = np.trunc(len(shuffled) * 0.64) # train
print ts
vs = np.trunc(len(shuffled) * 0.16)# valid
print vs
tts = np.trunc(len(shuffled) * 0.20 )# test
print tts
train_input, train_label = get_input_label(shuffled[:ts])
valid_input, valid_label = get_input_label(shuffled[ts:ts+vs])
test_input, test_label = get_input_label(shuffled[ts+vs:])
return train_input, train_label, valid_input, valid_label, test_input, test_label
def sort_by_cells(self):
zp, yp, xp = self.zp, self.yp, self.xp
vz, vy, vx = self.vz, self.vy, self.vx
Lz, Ly, Lx = self.Lz, self.Ly, self.Lx
cell = self.cell
cell_span = self.cell_span
N_cells = self.N_cells
Cz = Lz/N_cells
Cy = Ly/N_cells
Cx = Lx/N_cells
zp_cell = np.trunc(zp/Cz)
yp_cell = np.trunc(yp/Cy)
xp_cell = np.trunc(xp/Cx)
cell[:] = xp_cell+yp_cell*N_cells+zp_cell*N_cells**2
s = cell.argsort()
zp[:] = zp[s]
yp[:] = yp[s]
xp[:] = xp[s]
vz[:] = vz[s]
vy[:] = vy[s]
vx[:] = vx[s]
cell[:] = cell[s]
cell_span[:-1] = np.searchsorted(cell, self.cell_vals)
cell_span[-1] = self.N
def time2hms(time_in_seconds):
temp = time_in_seconds/3600.0
hours = np.trunc(temp)
minutes = np.trunc((temp - hours)*60)
seconds = (temp - hours - minutes/60)*3600
return (hours, minutes, seconds)
def lim_precision_inputs(inputs, bits, precision):
lim_inputs = np.empty(inputs.shape)
for index, value in np.ndenumerate(inputs):
if value>=0:
lim_inputs[index] = np.trunc((value * (1 << precision)) % (1 << bits) + 0.5)# * (1.0 / (1 << precision))
else:
lim_inputs[index] = -np.trunc((-value * (1 << precision)) % (1 << bits) + 0.5)# * (1.0 / (1 << precision))
return lim_inputs.astype(theano.config.floatX)
def plot(self, database, dsid):
"""Plot positions of all counterparts for all (unique) sources for
the given dataset.
The positions of all (unique) sources in the running catalog are
at the centre, whereas the positions of all their associated
sources are scattered around the central point. Axes are in
arcsec relative to the running catalog position.
"""
query = """\
SELECT x.id
,x.ra
,x.decl
,3600 * (x.ra - r.wm_ra) as ra_dist_arcsec
,3600 * (x.decl - r.wm_decl) as decl_dist_arcsec
,x.ra_err/2
,x.decl_err/2
,r.wm_ra_err/2
,r.wm_decl_err/2
FROM assocxtrsource a
,extractedsource x
,runningcatalog r
,image im1
WHERE a.runcat = r.id
AND a.xtrsrc = x.id
AND x.image = im1.id
AND im1.dataset = %s
"""
results = zip(*database.db.get(query, dsid))
if not results:
return None
xtrsrc_id = results[0]
ra = results[1]
decl = results[2]
ra_dist_arcsec = results[3]
decl_dist_arcsec = results[4]
ra_err = results[5]
decl_err = results[6]
wm_ra_err = results[7]
wm_decl_err = results[8]
axes = self.figure.add_subplot(1, 1, 1)
axes.errorbar(ra_dist_arcsec, decl_dist_arcsec, xerr=ra_err, yerr=decl_err,
fmt='+', color='b', label="xtr")
axes.set_xlabel(r'RA (arcsec)')
axes.set_ylabel(r'DEC (arcsec)')
lim = 1 + max(int(numpy.trunc(max(abs(min(ra_dist_arcsec)),
abs(max(ra_dist_arcsec))))),
int(numpy.trunc(max(abs(min(decl_dist_arcsec)),
abs(max(decl_dist_arcsec))))))
axes.set_xlim(xmin= -lim, xmax=lim)
axes.set_ylim(ymin= -lim, ymax=lim)
axes.grid(False)
# Shifts plot spacing to ensure that axes labels are displayed
self.figure.tight_layout()
def bprop_input (self, input, output):
sx = None
if self.thresh != 0:
sx = input.x / self.thresh
np.trunc(sx, sx)
np.sign(sx, sx)
else:
sx = np.sign(input.x)
input.dx += sx * output.dx
def trunc(x):
"""
Truncate the values to the integer value without rounding
"""
if isinstance(x, UncertainFunction):
mcpts = np.trunc(x._mcpts)
return UncertainFunction(mcpts)
else:
return np.trunc(x)
def gather(data,lc):
i = numpy.trunc(lc[0])
j = numpy.trunc(lc[1])
di = lc[0] - i
dj = lc[1] - j
return (data[i][j]*(1-di)*(1-dj) +
data[i+1][j]*(di)*(1-dj) +
data[i][j+1]*(1-di)*(dj) +
data[i+1][j+1]*(di)*(dj))
def truncate(self, h, m, s):
d = np.trunc(s / self.slim)
s = s % self.slim
m += d
d = np.trunc(m / self.mlim)
m = m % self.mlim
h += d
d = np.trunc(h / self.hlim)
h = h % self.hlim
return [int(h), int(m), float(s)]
def scatter(data,lc,value):
i = numpy.trunc(lc[0])
j = numpy.trunc(lc[1])
di = lc[0] - i
dj = lc[1] - j
data[i][j] += (1-di)*(1-dj)*value
data[i+1][j] += (di)*(1-dj)*value
data[i][j+1] += (1-di)*(dj)*value
data[i+1][j+1] += (di)*(dj)*value
def julian_date(yr, mo, d, hr, minute, sec, leap_sec=False):
x = (7*(yr + np.trunc((mo + 9)/12)))/4.0
y = (275*mo)/9.0
if leap_sec:
t = 61.0
else:
t = 60.0
z = (sec/t + minute)/60.0 + hr
jd = 367*yr - np.trunc(x) + np.trunc(y) + d + 1721013.5 + z/24.0
return jd
def tens_nort(drnms, labels,
mza = 100.,
mzb = 1000.,
mz_step = 1e-0, # 0.01
blcleaned = True):
ulist = []
llist = []
elist = []
for drnm in drnms:
if blcleaned:
fnm = drnm + '_blcleaned.npz'
else:
fnm = drnm + '.npz'
f = np.load("./" + drnm + '/' + fnm)
negsp = f['negsp']
negmz = f['negmz']
possp = f['possp']
posmz = f['posmz']
label = f['label']
for k in xrange(negsp.shape[1]): # must be equal to pos_s.shape[1] // number of samples
if label[k] not in labels:
continue
llist.append(label[k])
elist.append(drnm)
# u is 2D array: (m/z, polarity) = (m/z, 2), but polarity is fixednow
u = np.zeros( ( (mzb - mza) / mz_step, 2))
for i in xrange(negsp.shape[0]):
if negsp[i, k] != 0:
if (negmz[i] < mza) or (negmz[i] >= mzb):
continue
mzind = np.trunc((negmz[i] - mza)/ mz_step)
mzind = int(mzind)
#if u[mzind, rtind, 0] != 0:
u[mzind, 0] = max(u[mzind, 0], negsp[i, k]) # to deal with crossing values
for i in xrange(possp.shape[0]):
if possp[i, k] != 0:
if (posmz[i] < mza) or (posmz[i] >= mzb):
continue
mzind = np.trunc((posmz[i] - mza)/ mz_step)
mzind = int(mzind)
#if u[mzind, rtind, 0] != 0:
u[mzind, 1] = max(u[mzind, 1], possp[i, k]) # to deal with crossing values
ulist.append(u)
u = np.array(ulist)
return u, llist, elist
def jd2gregorian(jd):
t1900 = (jd - 2415019.5)/365.25
year = 1900 + np.trunc(t1900)
leap_years = np.trunc((year - 1900 - 1)*0.25)
days = (jd - 2415019.5) - ((year - 1900)*365.0 + leap_years)
if days < 1.0:
year = year - 1
leap_years = np.trunc((year - 1900 - 1)*0.25)
days = (jd - 2415019.5) - ((year - 1900)*365.0 + leap_years)
(month, day, hours, minutes, seconds) = days2ymdhms(days, year)
return (year, month, day, hours, minutes, seconds)
def base_link_coords_lethal(self, origin, yaw, x, y,):
x_rot = x*np.cos(yaw) - y*np.sin(yaw)
y_rot = y*np.cos(yaw) + x*np.sin(yaw)
x_index = np.trunc(origin[1] + x_rot/self.costmap_info.resolution).astype('i2')
y_index = np.trunc(origin[0] + y_rot/self.costmap_info.resolution).astype('i2')
if ( x_index < self.costmap_info.height and
y_index < self.costmap_info.height and
x_index >= 0 and
y_index >= 0 ):
return self.costmap[y_index, x_index] > self._lethal_threshold
else:
rospy.logwarn("SEARCH_AREA_CHECK off costmap! Returning True.")
return True
def interpolate_at_equidistant_time_steps(self):
#determine min and max of time series
min_t=numpy.min(self.ta)/RESAMPLE_TIME_STEP
max_t=numpy.max(self.ta)/RESAMPLE_TIME_STEP
# mapping to quantization steps RESAMPLE_TIME_STEP
min_ti=numpy.trunc(min_t)*RESAMPLE_TIME_STEP
max_ti=numpy.trunc(max_t)*RESAMPLE_TIME_STEP
number_of_samples=numpy.rint((max_ti-min_ti)/RESAMPLE_TIME_STEP+1+EPSILON)
#print number_of_samples
self.edta=numpy.linspace(min_ti, max_ti, number_of_samples)
#print self.edta
#scipy.interpolate.splrep: Find the B-spline representation of 1-D curve.
rep = scipy.interpolate.splrep(self.ta,self.ca, k=1)
self.edca=interpolate.splev(self.edta,rep)
def _update_fields(self):
from numpy import trunc
# map mouse location to array index
frotmp = int(trunc(self.custtool.yval))
totmp = int(trunc(self.custtool.xval))
# check if within range
sh = self.matrix_data_ref[self.edge_parameter_name].shape
# assume matrix whose shape is (# of rows, # of columns)
if frotmp >= 0 and frotmp < sh[0] and totmp >= 0 and totmp < sh[1]:
self.fro = self.labels[frotmp]
self.to = self.labels[totmp]
self.val = self.matrix_data_ref[self.edge_parameter_name][frotmp, totmp]
def _update_fields(self):
from numpy import trunc
# map mouse location to array index
frotmp = int(trunc(self.custtool.yval))
totmp = int(trunc(self.custtool.xval))
# check if within range
sh = self.data[self.data_name].shape
# assume matrix whose shape is (# of rows, # of columns)
if frotmp >= 0 and frotmp < sh[0] and totmp >= 0 and totmp < sh[1]:
self.fro = frotmp
self.to = totmp
self.val = self.data[self.data_name][self.fro, self.to]
def gyf2gms(gyf):
"""Conversion (g.frac) --> (gg,mm,ss.sssss)
Por ejemplo:
gyf2gms(-61.5) = (-61,30,0.0)
gyf2gms(-0.75) = (0,-45,0.0)"""
# hacer las cuentas
g = trunc(gyf)
m = (gyf - g) * 60
s = (m - trunc(m)) * 60
# preparar la presentación
g = int(g)
m = int(m) if g == 0 else int(abs(m))
s = abs(s) if m != 0 and g != 0 else s
return (g, m, s)
def line_blocked(self, userdata):
if self.debug_map_pub.get_num_connections() > 0:
publish_debug = True
else:
publish_debug = False
costmap = self.costmap
lethal_threshold = 90
blocked_threshold = 0.30
check_width = userdata.blocked_check_width/2
check_dist = userdata.blocked_check_distance
resolution = costmap.info.resolution
origin = (np.trunc(costmap.info.width/2),
np.trunc(costmap.info.height/2))
yaw = userdata.line_yaw
ll = self.check_point(origin, check_width, (yaw - math.pi/2), resolution)
ul = self.check_point(origin, check_width, (yaw + math.pi/2), resolution)
lr = self.check_point(ll[0], check_dist, yaw, resolution)
ur = self.check_point(ul[0], check_dist, yaw, resolution)
map_np = np.array(costmap.data, dtype='i1').reshape((costmap.info.height,costmap.info.width))
start_points = bresenham.points(ll, ul)
end_points = bresenham.points(lr, ur)
total_count = len(start_points)
#check lines for lethal values
blocked_count = 0
for start, end in zip(start_points, end_points):
line = bresenham.points(start[None,:], end[None,:])
line_vals = map_np[line[:,1], line[:,0]]
max_val = (np.amax(line_vals))
if np.any(line_vals > lethal_threshold):
blocked_count += 1
#for debug, mark lethal lines
if publish_debug: map_np[line[:,1], line[:,0]] = 64
#if anything is subscribing to the test map, publish it
if publish_debug:
map_np[start_points[:,1], start_points[:,0]] = 64
map_np[end_points[:,1], end_points[:,0]] = 64
costmap.data = list(np.reshape(map_np, -1))
self.debug_map_pub.publish(costmap)
if (blocked_count/float(total_count)) > blocked_threshold:
return True
return False
def contruct_classes(self, trn_idxs, corpus_mtrx_lst, cls_gnr_tgs, bagging_param=None):
inds_per_gnr = dict()
# inds = list()
last_gnr_tag = None
for gnr_tag in np.unique(cls_gnr_tgs[trn_idxs]):
inds_per_gnr[self.genres_lst[gnr_tag - 1]] = trn_idxs[
np.where(cls_gnr_tgs[trn_idxs] == gnr_tag)[0]
]
gnr_classes = dict()
for g, inds in inds_per_gnr.items():
if self.bagging and bagging_param:
# # # # # # #
shuffled_train_idxs = np.random.permutation(inds)
# print shuffled_train_idxs
# keep bagging_parram percent
bg_trn_ptg = int(np.trunc(shuffled_train_idxs.size * bagging_param))
# print bg_trn_ptg
bag_idxs = shuffled_train_idxs[0:bg_trn_ptg]
# print bag_idxs
elif not self.bagging and not bagging_param:
bag_idxs = inds
else:
raise Exception(
'contruct_classes(): Bagging triggerd with not bagging_param argument'
)
# Merge All Term-Frequency Dictionaries created by the Raw Texts
gnr_classes[g] = corpus_mtrx_lst[self.gnrlst_idx[g]][bag_idxs, :].mean(axis=0)
return gnr_classes
def randomPairs(n_records, sample_size):
"""
Return random combinations of indices for a square matrix of size
n records
"""
if n_records < 2 :
raise ValueError("Needs at least two records")
n = n_records * (n_records - 1) / 2
# numpy doesn't always throw an overflow error so we need to
# check to make sure that the largest number we'll use is smaller
# than the numpy's maximum unsigned integer
if 8 * n > numpy.iinfo('uint').max :
return randomPairsWithReplacement(n_records, sample_size)
if sample_size >= n:
if sample_size > n :
warnings.warn("Requested sample of size %d, only returning %d possible pairs" % (sample_size, n))
random_indices = numpy.arange(n)
else :
random_indices = numpy.random.randint(int(n), size=sample_size)
random_indices = random_indices.astype('uint')
b = 1 - 2 * n_records
x = numpy.trunc((-b - numpy.sqrt(b ** 2 - 8 * random_indices)) / 2)
y = random_indices + x * (b + x + 2) / 2 + 1
stacked = numpy.column_stack((x, y)).astype(int)
return [(p.item(), q.item()) for p, q in stacked]
def transfCartToCol2(m, linx, liny, linu, linv):
start = time.time()
lu, lv = len(linu), len(linv)
print "lu, lv:", lu, lv
minx, maxx = linx[0], linx[-1]
miny, maxy = liny[0], liny[-1]
dx = 1/abs(linx[1]-linx[0])
#print minx, miny, maxx, maxy
U, V = np.meshgrid(linu, linv)
coord = np.zeros((lv, lu, 2))
coord[:,:,:] = fromColi(U,V)
coord[:,:,0] = (np.ceil(dx*coord[:,:,0])-1)/dx
## transforme: -0.8 en -1, et 0.8 en 0: utile par car on les x<0 n'existe pas
#(si on transforme -0.8 en 0, on n'echoue la selection)
coord[:,:,1] = (np.trunc(dx*coord[:,:,1]))/dx
## transforme: -0.8 en 0, et 0.8 en 0: utile pour ne pas depasser le y max
mt = np.zeros((lv, lu))
for u in xrange(0, lu):
for v in xrange(0, lv):
x, y = coord[v,u,0], coord[v, u, 1]
if ((x > (maxx)) | (x < (minx)) | (y > (maxy)) | (y < (miny)) ):
mt[v,u] = 0.0
else:
ind_x = np.nonzero(linx==x)[0][0]
ind_y = np.nonzero(liny==y)[0][0]
mt[v, u] = m[ind_y,ind_x]#np.sum(m[ind_y-1:ind_y+1,ind_x-1:ind_x+1])/9
conn_time = time.time()-start
a = time.gmtime(conn_time)
print time.strftime("\ntask time for map transformation: %H:%M:%S",a )
return(mt)
def transfColToCart(m, linx, liny, linu, linv):
##picture tranformation
start = time.time()
lx, ly = len(linx), len(liny)
#print "lu, lv:", lx, ly
minu, maxu = linu[0], linu[-1]
minv, maxv = linv[0], linv[-1]
du = 1/abs(linu[1]-linu[0])
#print minx, miny, maxx, maxy
X, Y = np.meshgrid(linx, liny)
coord = np.zeros((ly, lx, 2))
coord[:,:,:] = toColi(X,Y)
coord[:,:,0] = (np.ceil(du*coord[:,:,0])-1)/du
coord[:,:,1] = (np.trunc(du*coord[:,:,1]))/du
coord[np.isnan(coord)] = 0
mt = np.zeros((ly, lx))
for x in xrange(0, lx):
for y in xrange(0, ly):
u, v = coord[y,x,0], coord[y, x, 1]
if ((u > (maxu)) | (u < (minu)) | (v > (maxv)) | (v < (minv)) ):
mt[y,x] = 0.0
else:
#print u,v
ind_u = np.nonzero(linu==u)[0][0]
ind_v = np.nonzero(linv==v)[0][0]
mt[y, x] = m[ind_v,ind_u]#np.sum(m[ind_y-1:ind_y+1,ind_x-1:ind_x+1])/9
conn_time = time.time()-start
a = time.gmtime(conn_time)
print time.strftime("\ntask time for map transformation: %H:%M:%S",a )
return(mt)
def signal(self, fs, atten, caldb, calv):
if self._filename is None:
# allow lack of file to not cause error, catch in GUI when necessary?
logger = logging.getLogger('main')
logger.warn('Vocalization signal request without a file')
return np.array([0,0])
if not self._findFile():
return np.array([0,0])
fs, wavdata = audioread(self._filename)
if fs != fs:
print 'specified', fs, 'wav file', fs
raise Exception("specified samplerate does not match wav stimulus")
#truncate to nears ms
duration = float(len(wavdata))/fs
# print 'duration {}, desired {}'.format(duration, np.trunc(duration*1000)/1000)
desired_npts = int((np.trunc(duration*1000)/1000)*fs)
# print 'npts. desired', len(wavdata), desired_npts
wavdata = wavdata[:desired_npts]
amp_scale = signal_amplitude(wavdata, fs)
signal = ((wavdata/amp_scale)*self.amplitude(caldb, calv))
if self._risefall > 0:
rf_npts = int(self._risefall * fs) / 2
wnd = hann(rf_npts*2) # cosine taper
signal[:rf_npts] = signal[:rf_npts] * wnd[:rf_npts]
signal[-rf_npts:] = signal[-rf_npts:] * wnd[rf_npts:]
return signal
def get_pointcloud2(self, grid, position, target_frame, transform, range=12.0):
header = std_msg.Header(0, rospy.Time.now(), target_frame)
grid_np = np.array(grid.data, dtype='u1').reshape((grid.info.height,grid.info.width))
origin = np.array((grid.info.origin.position.x,
grid.info.origin.position.y,
grid.info.origin.position.z))
world2map = lambda x:np.clip(trunc((x-origin)/grid.info.resolution),
zeros((3,)),
np.r_[grid.info.width-1, grid.info.height-1,0])
ll = world2map(np.r_[position.x-range, position.y-range, 0])
ur = world2map(np.r_[position.x+range, position.y+range, 0])
submap = grid_np[ll[1]:ur[1],ll[0]:ur[0]]
mappts = np.c_[np.where(submap==100)][:,::-1] + ll[:2]
map2world = lambda x:(x+0.5)*grid.info.resolution+origin[:2]
wpts = np.array([map2world(x) for x in mappts])
pcpts=[]
if len(wpts>0):
for z in np.linspace(0,1,11):
pcpts.append(np.c_[wpts,z*ones_like(wpts[:,0])])
pcpts = np.vstack(pcpts)
pcpts = np.einsum("ki,...ji", transform, np.c_[pcpts,np.ones((len(pcpts),1))])[:,:3]
else:
pcpts = []
pc = pc2.create_cloud_xyz32(header, pcpts)
return pc
def randomPairs(n_records, sample_size, zero_indexed=True):
"""
Return random combinations of indices for a square matrix of size
n records
"""
if n_records < 2:
raise ValueError("Needs at least two records")
n = n_records * (n_records - 1) / 2
if sample_size >= n:
random_indices = numpy.arange(n)
numpy.random.shuffle(random_indices)
else:
try:
random_indices = numpy.array(random.sample(xrange(n), sample_size))
except OverflowError:
# If the population is very large relative to the sample
# size than we'll get very few duplicates by chance
logging.warning("There may be duplicates in the sample")
sample = numpy.array([random.sample(xrange(n_records), 2) for _ in xrange(sample_size)])
return numpy.sort(sample, axis=1)
b = 1 - 2 * n_records
x = numpy.trunc((-b - numpy.sqrt(b ** 2 - 8 * random_indices)) / 2)
y = random_indices + x * (b + x + 2) / 2 + 1
if not zero_indexed:
x += 1
y += 1
return numpy.column_stack((x, y)).astype(int)
def gumr(xn, xu):
z2 = np.trunc(np.log10(xu))+1
z1 = np.around(xu/(10**z2), 3)
y1 = np.around(xn*10**(-z2), 2)
value = y1*10**z2
uncert = z1*10**z2
return('%g'%value, '%g'%uncert)
def ImColorNamingTSELabDescriptor(ima, positions=None, patchSize=1):
# Constants
numColors = 11 # Number of colors
numAchromatics = 3 # Number of achromatic colors
numChromatics = numColors - numAchromatics # Number of chromatic colors
# Initializations
numLevels = np.size(thrL) - 1 # Number of Lightness levels in the model
# Image conversion: sRGB to CIELab
Lab = RGB2Lab(ima)
if positions != None:
if patchSize == 1:
Lab = Lab[positions[:, 0], positions[:, 1], :].reshape((1, -1, 3))
else:
LabPatch = np.zeros(
(positions.shape[0], (2 * np.trunc(patchSize / 2) + 1)**2, 3))
padSz = (int(np.trunc(patchSize / 2)),
int(np.trunc(patchSize / 2)))
Lab = np.pad(Lab, (padSz, padSz, (0, 0)), 'symmetric')
positions += padSz[0]
c = 0
for x in range(-padSz[0], padSz[0] + 1):
for y in range(-padSz[0], padSz[0] + 1):
LabPatch[:, c, :] = Lab[positions[:, 0] + y,
positions[:, 1] + x, :]
c += 1
Lab = LabPatch
S = np.shape(Lab)
if Lab.ndim == 2:
L = Lab[:, 0].flatten()
a = Lab[:, 1].flatten()
b = Lab[:, 2].flatten()
nr = S[0]
nc = 1
# Image dimensions: rows, columns, and channels
else:
L = Lab[:, :, 0].flatten()
a = Lab[:, :, 1].flatten()
b = Lab[:, :, 2].flatten()
nr = S[0]
nc = S[1]
# Image dimensions: rows, columns, and channels
npix = nr * nc # Number of pixels
CD = np.zeros((npix, numColors)) # Color descriptor to store results
# Assignment of the sample to its corresponding level
m = np.zeros(np.shape(L))
m[np.where(L == 0)[0]] = 1 # Pixels with L=0 assigned to level 1
for k in range(1, numLevels + 1):
m = m + np.double(thrL[k - 1] < L) * np.double(
L <= thrL[k]) * np.double(k)
m = m.astype(int) - 1
# Computing membership values to chromatic categories
for k in range(numChromatics):
tx = np.reshape(parameters[k, 0, m], (npix, 1))
ty = np.reshape(parameters[k, 1, m], (npix, 1))
alfa_x = np.reshape(parameters[k, 2, m], (npix, 1))
alfa_y = np.reshape(parameters[k, 3, m], (npix, 1))
beta_x = np.reshape(parameters[k, 4, m], (npix, 1))
beta_y = np.reshape(parameters[k, 5, m], (npix, 1))
beta_e = np.reshape(parameters[k, 6, m], (npix, 1))
ex = np.reshape(parameters[k, 7, m], (npix, 1))
ey = np.reshape(parameters[k, 8, m], (npix, 1))
angle_e = np.reshape(parameters[k, 9, m], (npix, 1))
#figure;plot(angle_e); show()
CD[:, k] = (np.double(beta_e != 0.0) *
TripleSigmoid_E(np.vstack((a, b)), tx, ty, alfa_x, alfa_y,
beta_x, beta_y, beta_e, ex, ey, angle_e)).T
# Computing membership values to achromatic categories
valueAchro = np.squeeze(
np.maximum(1.0 - np.reshape(np.sum(CD, axis=1), (npix, 1)),
np.zeros((npix, 1))))
CD[:, numChromatics +
0] = valueAchro * Sigmoid(L, paramsAchro[0, 0], paramsAchro[0, 1])
CD[:, numChromatics + 1] = valueAchro * Sigmoid(
L, paramsAchro[1, 0], paramsAchro[1, 1]) * Sigmoid(
L, paramsAchro[2, 0], paramsAchro[2, 1])
CD[:, numChromatics +
2] = valueAchro * Sigmoid(L, paramsAchro[3, 0], paramsAchro[3, 1])
# Color descriptor with color memberships to all the categories (one color in each plane)
if positions == None or patchSize > 1:
CD = np.reshape(CD, (nr, nc, numColors))
if patchSize > 1:
CD = np.sum(CD, axis=1)
CD = CD / np.tile(np.sum(CD, axis=1).reshape(-1, 1), (1, numColors))
if Lab.ndim == 2:
CD = np.reshape(CD, (-1, CD.shape[2]))
CD = CD / np.expand_dims(np.sum(CD, axis=len(CD.shape) - 1),
axis=len(CD.shape) - 1)
return CD
# %% # 36. 用五种不同的方法去提取一个随机数组的整数部分(★★☆) z = np.random.uniform(-10, 10, 10) # 1.z - z%1 (向下取整) print(z) print(z - z % 1) # 2.around (四舍五入) print(np.around(z, 0)) # 3.ceil(向上取整) print(np.ceil(z)) # 4.floor(向下取整) print(np.floor(z)) # 5.trunc(向0取整) print(np.trunc(z)) # 6.rint 四舍五入 print(np.rint(z)) # 7.astype print(z.astype(int)) # %% '''37. 创建一个5x5的矩阵，其中每行的数值范围从0到4 (★★☆)''' """ 当矩阵 x 非常大时，在Python中计算显式循环可能会很慢。注意， 向矩阵 x 的每一行添加向量 v 等同于通过垂直堆叠多个 v 副本来形成矩阵 vv， 然后执行元素的求和x 和 vv。 我们可以像如下这样实现这种方法： """ """ 如果数组不具有相同的rank，则将较低等级数组的形状添加1， 直到两个形状具有相同的长度。
def get_tax_of_payment_transportation(self, engineer_history):
if engineer_history:
return numpy.trunc((self.payment_transportation or 0) * engineer_history.payment_tax.rate_if_exclude_tax)
else:
return 0
def dftregistration(buf1ft,
buf2ft,
usfac=1,
return_registered=False,
return_error=False,
zeromean=True,
DEBUG=False,
maxoff=None,
nthreads=1,
use_numpy_fft=False):
"""
translated from matlab:
http://www.mathworks.com/matlabcentral/fileexchange/18401-efficient-subpixel-image-registration-by-cross-correlation/content/html/efficient_subpixel_registration.html
Efficient subpixel image registration by crosscorrelation. This code
gives the same precision as the FFT upsampled cross correlation in a
small fraction of the computation time and with reduced memory
requirements. It obtains an initial estimate of the crosscorrelation peak
by an FFT and then refines the shift estimation by upsampling the DFT
only in a small neighborhood of that estimate by means of a
matrix-multiply DFT. With this procedure all the image points are used to
compute the upsampled crosscorrelation.
Manuel Guizar - Dec 13, 2007
Portions of this code were taken from code written by Ann M. Kowalczyk
and James R. Fienup.
J.R. Fienup and A.M. Kowalczyk, "Phase retrieval for a complex-valued
object by using a low-resolution image," J. Opt. Soc. Am. A 7, 450-458
(1990).
Citation for this algorithm:
Manuel Guizar-Sicairos, Samuel T. Thurman, and James R. Fienup,
"Efficient subpixel image registration algorithms," Opt. Lett. 33,
156-158 (2008).
Inputs
buf1ft Fourier transform of reference image,
DC in (1,1) [DO NOT FFTSHIFT]
buf2ft Fourier transform of image to register,
DC in (1,1) [DO NOT FFTSHIFT]
usfac Upsampling factor (integer). Images will be registered to
within 1/usfac of a pixel. For example usfac = 20 means the
images will be registered within 1/20 of a pixel. (default = 1)
Outputs
output = [error,diffphase,net_row_shift,net_col_shift]
error Translation invariant normalized RMS error between f and g
diffphase Global phase difference between the two images (should be
zero if images are non-negative).
net_row_shift net_col_shift Pixel shifts between images
Greg (Optional) Fourier transform of registered version of buf2ft,
the global phase difference is compensated for.
"""
# this function is translated from matlab, so I'm just going to pretend
# it is matlab/pylab
from numpy import conj, abs, arctan2, sqrt, real, imag, shape, zeros, trunc, ceil, floor, fix
from numpy.fft import fftshift, ifftshift
fft2, ifft2 = fftn, ifftn = fast_ffts.get_ffts(nthreads=nthreads,
use_numpy_fft=use_numpy_fft)
# Compute error for no pixel shift
if usfac == 0:
raise ValueError("Upsample Factor must be >= 1")
CCmax = sum(sum(buf1ft * conj(buf2ft)))
rfzero = sum(abs(buf1ft)**2)
rgzero = sum(abs(buf2ft)**2)
error = 1.0 - CCmax * conj(CCmax) / (rgzero * rfzero)
error = sqrt(abs(error))
diffphase = arctan2(imag(CCmax), real(CCmax))
output = [error, diffphase]
# Whole-pixel shift - Compute crosscorrelation by an IFFT and locate the
# peak
elif usfac == 1:
[m, n] = shape(buf1ft)
CC = ifft2(buf1ft * conj(buf2ft))
if maxoff is None:
rloc, cloc = np.unravel_index(abs(CC).argmax(), CC.shape)
CCmax = CC[rloc, cloc]
else:
# set the interior of the shifted array to zero
# (i.e., ignore it)
CC[maxoff:-maxoff, :] = 0
CC[:, maxoff:-maxoff] = 0
rloc, cloc = np.unravel_index(abs(CC).argmax(), CC.shape)
CCmax = CC[rloc, cloc]
rfzero = sum(abs(buf1ft)**2) / (m * n)
rgzero = sum(abs(buf2ft)**2) / (m * n)
error = 1.0 - CCmax * conj(CCmax) / (rgzero * rfzero)
error = sqrt(abs(error))
diffphase = arctan2(imag(CCmax), real(CCmax))
md2 = fix(m / 2)
nd2 = fix(n / 2)
if rloc > md2:
row_shift = rloc - m
else:
row_shift = rloc
if cloc > nd2:
col_shift = cloc - n
else:
col_shift = cloc
#output=[error,diffphase,row_shift,col_shift];
output = [row_shift, col_shift]
# Partial-pixel shift
else:
if DEBUG: import pylab
# First upsample by a factor of 2 to obtain initial estimate
# Embed Fourier data in a 2x larger array
[m, n] = shape(buf1ft)
mlarge = m * 2
nlarge = n * 2
CClarge = zeros([mlarge, nlarge], dtype='complex')
#CClarge[m-fix(m/2):m+fix((m-1)/2)+1,n-fix(n/2):n+fix((n-1)/2)+1] = fftshift(buf1ft) * conj(fftshift(buf2ft));
CClarge[round(mlarge / 4.):round(mlarge / 4. * 3),
round(nlarge /
4.):round(nlarge / 4. *
3)] = fftshift(buf1ft) * conj(fftshift(buf2ft))
# note that matlab uses fix which is trunc... ?
# Compute crosscorrelation and locate the peak
CC = ifft2(ifftshift(CClarge))
# Calculate cross-correlation
if maxoff is None:
rloc, cloc = np.unravel_index(abs(CC).argmax(), CC.shape)
CCmax = CC[rloc, cloc]
else:
# set the interior of the shifted array to zero
# (i.e., ignore it)
CC[maxoff:-maxoff, :] = 0
CC[:, maxoff:-maxoff] = 0
rloc, cloc = np.unravel_index(abs(CC).argmax(), CC.shape)
CCmax = CC[rloc, cloc]
if DEBUG:
pylab.figure(1)
pylab.clf()
pylab.subplot(131)
pylab.imshow(real(CC))
pylab.title("Cross-Correlation (upsampled 2x)")
pylab.subplot(132)
ups = dftups((buf1ft) * conj((buf2ft)), mlarge, nlarge, 2, 0, 0)
pylab.title("dftups upsampled 2x")
pylab.imshow(real(((ups))))
pylab.subplot(133)
pylab.imshow(real(CC) / real(ups))
pylab.title("Ratio upsampled/dftupsampled")
print "Upsample by 2 peak: ", rloc, cloc, " using dft version: ", np.unravel_index(
abs(ups).argmax(), ups.shape)
#print np.unravel_index(ups.argmax(),ups.shape)
# Obtain shift in original pixel grid from the position of the
# crosscorrelation peak
[m, n] = shape(CC)
md2 = trunc(m / 2)
nd2 = trunc(n / 2)
if rloc > md2:
row_shift2 = rloc - m
else:
row_shift2 = rloc
if cloc > nd2:
col_shift2 = cloc - n
else:
col_shift2 = cloc
row_shift2 = row_shift2 / 2.
col_shift2 = col_shift2 / 2.
if DEBUG:
print "row_shift/col_shift from ups2: ", row_shift2, col_shift2
# If upsampling > 2, then refine estimate with matrix multiply DFT
if usfac > 2:
#%% DFT computation %%%
# Initial shift estimate in upsampled grid
zoom_factor = 1.5
if DEBUG: print row_shift2, col_shift2
row_shift0 = round(row_shift2 * usfac) / usfac
col_shift0 = round(col_shift2 * usfac) / usfac
dftshift = trunc(ceil(usfac * zoom_factor) / 2)
#% Center of output array at dftshift+1
if DEBUG:
print 'dftshift,rs,cs,zf:', dftshift, row_shift0, col_shift0, usfac * zoom_factor
# Matrix multiply DFT around the current shift estimate
roff = dftshift - row_shift0 * usfac
coff = dftshift - col_shift0 * usfac
upsampled = dftups((buf2ft * conj(buf1ft)),
ceil(usfac * zoom_factor),
ceil(usfac * zoom_factor), usfac, roff, coff)
#CC = conj(dftups(buf2ft.*conj(buf1ft),ceil(usfac*1.5),ceil(usfac*1.5),usfac,...
# dftshift-row_shift*usfac,dftshift-col_shift*usfac))/(md2*nd2*usfac^2);
CC = conj(upsampled) / (md2 * nd2 * usfac**2)
if DEBUG:
pylab.figure(2)
pylab.clf()
pylab.subplot(221)
pylab.imshow(abs(upsampled))
pylab.title('upsampled')
pylab.subplot(222)
pylab.imshow(abs(CC))
pylab.title('CC upsampled')
pylab.subplot(223)
pylab.imshow(
np.abs(np.fft.fftshift(np.fft.ifft2(buf2ft *
conj(buf1ft)))))
pylab.title('xc')
yy, xx = np.indices([m * usfac, n * usfac], dtype='float')
pylab.contour(
yy / usfac / 2. - 0.5 + 1, xx / usfac / 2. - 0.5 - 1,
np.abs(
dftups((buf2ft * conj(buf1ft)), m * usfac, n * usfac,
usfac)))
pylab.subplot(224)
pylab.imshow(
np.abs(
dftups((buf2ft * conj(buf1ft)),
ceil(usfac * zoom_factor),
ceil(usfac * zoom_factor), usfac)))
pylab.title('unshifted ups')
# Locate maximum and map back to original pixel grid
rloc, cloc = np.unravel_index(abs(CC).argmax(), CC.shape)
rloc0, cloc0 = np.unravel_index(abs(CC).argmax(), CC.shape)
CCmax = CC[rloc, cloc]
#[max1,loc1] = CC.max(axis=0), CC.argmax(axis=0)
#[max2,loc2] = max1.max(),max1.argmax()
#rloc=loc1[loc2];
#cloc=loc2;
#CCmax = CC[rloc,cloc];
rg00 = dftups(buf1ft * conj(buf1ft), 1, 1,
usfac) / (md2 * nd2 * usfac**2)
rf00 = dftups(buf2ft * conj(buf2ft), 1, 1,
usfac) / (md2 * nd2 * usfac**2)
#if DEBUG: print rloc,row_shift,cloc,col_shift,dftshift
rloc = rloc - dftshift #+ 1 # +1 # questionable/failed hack + 1;
cloc = cloc - dftshift #+ 1 # -1 # questionable/failed hack - 1;
#if DEBUG: print rloc,row_shift,cloc,col_shift,dftshift
row_shift = row_shift0 + rloc / usfac
col_shift = col_shift0 + cloc / usfac
#if DEBUG: print rloc/usfac,row_shift,cloc/usfac,col_shift
if DEBUG:
print "Off by: ", (0.25 - float(rloc) / usfac) * usfac, (
-0.25 - float(cloc) / usfac) * usfac
if DEBUG: print "correction was: ", rloc / usfac, cloc / usfac
if DEBUG:
print "Coordinate went from", row_shift2, col_shift2, "to", row_shift0, col_shift0, "to", row_shift, col_shift
if DEBUG: print "dftsh - usfac:", dftshift - usfac
if DEBUG:
print rloc, cloc, row_shift, col_shift, CCmax, dftshift, rloc0, cloc0
# If upsampling = 2, no additional pixel shift refinement
else:
rg00 = sum(sum(buf1ft * conj(buf1ft))) / m / n
rf00 = sum(sum(buf2ft * conj(buf2ft))) / m / n
row_shift = row_shift2
col_shift = col_shift2
error = 1.0 - CCmax * conj(CCmax) / (rg00 * rf00)
error = sqrt(abs(error))
diffphase = arctan2(imag(CCmax), real(CCmax))
# If its only one row or column the shift along that dimension has no
# effect. We set to zero.
if md2 == 1:
row_shift = 0
if nd2 == 1:
col_shift = 0
#output=[error,diffphase,row_shift,col_shift];
output = [row_shift, col_shift]
if return_error:
# simple estimate of the precision of the fft approach
output += [1. / usfac, 1. / usfac]
# Compute registered version of buf2ft
if (return_registered):
if (usfac > 0):
nr, nc = shape(buf2ft)
Nr = np.fft.ifftshift(
np.linspace(-np.fix(nr / 2),
np.ceil(nr / 2) - 1, nr))
Nc = np.fft.ifftshift(
np.linspace(-np.fix(nc / 2),
np.ceil(nc / 2) - 1, nc))
[Nc, Nr] = np.meshgrid(Nc, Nr)
Greg = buf2ft * np.exp(
1j * 2 * np.pi * (-row_shift * Nr / nr - col_shift * Nc / nc))
Greg = Greg * np.exp(1j * diffphase)
elif (usfac == 0):
Greg = buf2ft * np.exp(1j * diffphase)
output.append(Greg)
return output
def hep(trange=['2017-03-27', '2017-03-28'],
datatype='omniflux',
level='l2',
suffix='',
get_support_data=False,
varformat=None,
varnames=[],
downloadonly=False,
notplot=False,
no_update=False,
uname=None,
passwd=None,
time_clip=False,
ror=True,
version=None):
"""
This function loads data from the HEP experiment from the Arase mission
Parameters:
trange : list of str
time range of interest [starttime, endtime] with the format
'YYYY-MM-DD','YYYY-MM-DD'] or to specify more or less than a day
['YYYY-MM-DD/hh:mm:ss','YYYY-MM-DD/hh:mm:ss']
datatype: str
Data type; Valid options:
level: str
Data level; Valid options:
suffix: str
The tplot variable names will be given this suffix. By default,
no suffix is added.
get_support_data: bool
Data with an attribute "VAR_TYPE" with a value of "support_data"
will be loaded into tplot. By default, only loads in data with a
"VAR_TYPE" attribute of "data".
varformat: str
The file variable formats to load into tplot. Wildcard character
"*" is accepted. By default, all variables are loaded in.
varnames: list of str
List of variable names to load (if not specified,
all data variables are loaded)
downloadonly: bool
Set this flag to download the CDF files, but not load them into
tplot variables
notplot: bool
Return the data in hash tables instead of creating tplot variables
no_update: bool
If set, only load data from your local cache
time_clip: bool
Time clip the variables to exactly the range specified in the trange keyword
ror: bool
If set, print PI info and rules of the road
version: str
Set this value to specify the version of cdf files (such as "v01_02", "v01_03", ...)
Returns:
List of tplot variables created.
"""
file_res = 3600. * 24
prefix = 'erg_hep_'+level+'_'
if level == 'l2':
pathformat = 'satellite/erg/hep/'+level+'/'+datatype + \
'/%Y/%m/erg_hep_'+level+'_'+datatype + '_%Y%m%d_'
if version is None:
pathformat += 'v??_??.cdf'
else:
pathformat += version + '.cdf'
if level == 'l3':
pathformat = 'satellite/erg/hep/'+level + \
'/pa/%Y/%m/erg_hep_'+level+'_pa_%Y%m%d_'
if version is None:
pathformat += 'v??_??.cdf'
else:
pathformat += version + '.cdf'
initial_notplot_flag = False
if notplot:
initial_notplot_flag = True
if ((level == 'l2') and (datatype == 'omniflux')) or (datatype == '3dflux') or (level == 'l3'):
# to avoid failure of creation plot variables (at store_data.py) of hep
notplot = True
loaded_data = load(pathformat=pathformat, trange=trange, level=level, datatype=datatype, file_res=file_res, prefix=prefix, suffix=suffix, get_support_data=get_support_data,
varformat=varformat, varnames=varnames, downloadonly=downloadonly, notplot=notplot, time_clip=time_clip, no_update=no_update, uname=uname, passwd=passwd, version=version)
if (len(loaded_data) > 0) and ror:
try:
if isinstance(loaded_data, list):
if downloadonly:
cdf_file = cdflib.CDF(loaded_data[-1])
gatt = cdf_file.globalattsget()
else:
gatt = get_data(loaded_data[-1], metadata=True)['CDF']['GATT']
elif isinstance(loaded_data, dict):
gatt = loaded_data[list(loaded_data.keys())[-1]]['CDF']['GATT']
# --- print PI info and rules of the road
print(' ')
print(
'**************************************************************************')
print(gatt["LOGICAL_SOURCE_DESCRIPTION"])
print('')
print('PI: ', gatt['PI_NAME'])
print("Affiliation: "+gatt["PI_AFFILIATION"])
print('')
print('- The rules of the road (RoR) common to the ERG project:')
print(
' https://ergsc.isee.nagoya-u.ac.jp/data_info/rules_of_the_road.shtml.en')
print(
'- RoR for HEP data: https://ergsc.isee.nagoya-u.ac.jp/mw/index.php/ErgSat/Hep')
if level == 'l3':
print(
'- RoR for MGF data: https://ergsc.isee.nagoya-u.ac.jp/mw/index.php/ErgSat/Mgf')
print('')
print('Contact: erg_hep_info at isee.nagoya-u.ac.jp')
print(
'**************************************************************************')
except:
print('printing PI info and rules of the road was failed')
if initial_notplot_flag or downloadonly:
return loaded_data
if isinstance(loaded_data, dict):
if (level == 'l2') and (datatype == 'omniflux'):
tplot_variables = []
if prefix + 'FEDO_L' + suffix in loaded_data:
v_vars_min = loaded_data[prefix + 'FEDO_L' + suffix]['v'][0]
v_vars_max = loaded_data[prefix + 'FEDO_L' + suffix]['v'][1]
# log average of energy bins
v_vars = np.power(
10., (np.log10(v_vars_min) + np.log10(v_vars_max)) / 2.)
store_data(prefix + 'FEDO_L' + suffix, data={'x': loaded_data[prefix + 'FEDO_L' + suffix]['x'],
'y': loaded_data[prefix + 'FEDO_L' + suffix]['y'],
'v': v_vars},
attr_dict={'CDF':loaded_data[prefix + 'FEDO_L' + suffix]['CDF']})
tplot_variables.append(prefix + 'FEDO_L' + suffix)
if prefix + 'FEDO_H' + suffix in loaded_data:
v_vars_min = loaded_data[prefix + 'FEDO_H' + suffix]['v'][0]
v_vars_max = loaded_data[prefix + 'FEDO_H' + suffix]['v'][1]
# log average of energy bins
v_vars = np.power(
10., (np.log10(v_vars_min) + np.log10(v_vars_max)) / 2.)
store_data(prefix + 'FEDO_H' + suffix, data={'x': loaded_data[prefix + 'FEDO_H' + suffix]['x'],
'y': loaded_data[prefix + 'FEDO_H' + suffix]['y'],
'v': v_vars},
attr_dict={'CDF':loaded_data[prefix + 'FEDO_H' + suffix]['CDF']})
tplot_variables.append(prefix + 'FEDO_H' + suffix)
# remove minus valuse of y array
if prefix + 'FEDO_L' + suffix in tplot_variables:
clip(prefix + 'FEDO_L' + suffix, 0., 1.0e+10)
if prefix + 'FEDO_H' + suffix in tplot_variables:
clip(prefix + 'FEDO_H' + suffix, 0., 1.0e+10)
# set spectrogram plot option
options(prefix + 'FEDO_L' + suffix, 'Spec', 1)
options(prefix + 'FEDO_H' + suffix, 'Spec', 1)
# set y axis to logscale
options(prefix + 'FEDO_L' + suffix, 'ylog', 1)
options(prefix + 'FEDO_H' + suffix, 'ylog', 1)
# set yrange
options(prefix + 'FEDO_L' + suffix, 'yrange', [3.0e+01, 2.0e+03])
options(prefix + 'FEDO_H' + suffix, 'yrange', [7.0e+01, 2.0e+03])
# set ytitle
options(prefix + 'FEDO_L' + suffix, 'ytitle',
'HEP-L\nomniflux\nLv2\nEnergy')
options(prefix + 'FEDO_H' + suffix, 'ytitle',
'HEP-H\nomniflux\nLv2\nEnergy')
# set ysubtitle
options(prefix + 'FEDO_L' + suffix, 'ysubtitle', '[keV]')
options(prefix + 'FEDO_H' + suffix, 'ysubtitle', '[keV]')
# set ylim
if prefix + 'FEDO_L' + suffix in tplot_variables:
ylim(prefix + 'FEDO_L' + suffix, 30, 1800)
if prefix + 'FEDO_H' + suffix in tplot_variables:
ylim(prefix + 'FEDO_H' + suffix, 500, 2048)
# set z axis to logscale
options(prefix + 'FEDO_L' + suffix, 'zlog', 1)
options(prefix + 'FEDO_H' + suffix, 'zlog', 1)
# set zrange
options(prefix + 'FEDO_L' + suffix, 'zrange', [1.0e-15, 1.0e+06])
options(prefix + 'FEDO_H' + suffix, 'zrange', [1.0e-10, 1.0e+5])
# set ztitle
options(prefix + 'FEDO_L' + suffix,
'ztitle', '[/cm^{2}-str-s-keV]')
options(prefix + 'FEDO_H' + suffix,
'ztitle', '[/cm^{2}-str-s-keV]')
# set zlim
if prefix + 'FEDO_L' + suffix in tplot_variables:
zlim(prefix + 'FEDO_L' + suffix, 1e+0, 1e+5)
if prefix + 'FEDO_H' + suffix in tplot_variables:
zlim(prefix + 'FEDO_H' + suffix, 1e+0, 1e+5)
# change colormap option
options(prefix + 'FEDO_L' + suffix, 'Colormap', 'jet')
options(prefix + 'FEDO_H' + suffix, 'Colormap', 'jet')
return tplot_variables
if (level == 'l2') and (datatype == '3dflux'):
tplot_variables = []
v2_array = [i for i in range(15)]
if prefix + 'FEDU_L' + suffix in loaded_data:
store_data(prefix + 'FEDU_L' + suffix, data={'x': loaded_data[prefix + 'FEDU_L' + suffix]['x'],
'y': loaded_data[prefix + 'FEDU_L' + suffix]['y'],
'v1': np.sqrt(loaded_data[prefix + 'FEDU_L' + suffix]['v'][0, :] *
loaded_data[prefix + 'FEDU_L' + suffix]['v'][1, :]), # geometric mean for 'v1'
'v2': v2_array},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_L' + suffix]['CDF']})
tplot_variables.append(prefix + 'FEDU_L' + suffix)
clip(prefix + 'FEDU_L' + suffix, -1.0e+10, 1.0e+10)
if prefix + 'FEDU_H' + suffix in loaded_data:
store_data(prefix + 'FEDU_H' + suffix, data={'x': loaded_data[prefix + 'FEDU_H' + suffix]['x'],
'y': loaded_data[prefix + 'FEDU_H' + suffix]['y'],
'v1': np.sqrt(loaded_data[prefix + 'FEDU_H' + suffix]['v'][0, :] *
loaded_data[prefix + 'FEDU_H' + suffix]['v'][1, :]), # geometric mean for 'v1'
'v2': v2_array},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_H' + suffix]['CDF']})
tplot_variables.append(prefix + 'FEDU_H' + suffix)
clip(prefix + 'FEDU_H' + suffix, -1.0e+10, 1.0e+10)
return tplot_variables
if level == 'l3': # implementation for level = 'l3'
tplot_variables = []
if prefix + 'FEDU_L' + suffix in loaded_data:
L_energy_array_ave = np.sqrt(loaded_data[prefix + 'FEDU_L' + suffix]['v1'][0, :] *
loaded_data[prefix + 'FEDU_L' + suffix]['v1'][1, :]) # geometric mean for 'v1'
# get energy [keV] array for ytitle options
L_energy_array = np.trunc(L_energy_array_ave).astype(int)
non_negative_y_array = np.where(
loaded_data[prefix + 'FEDU_L' + suffix]['y'] < 0., np.nan, loaded_data[prefix + 'FEDU_L' + suffix]['y'])
store_data(prefix + 'FEDU_L' + suffix, data={'x': loaded_data[prefix + 'FEDU_L' + suffix]['x'],
'y': non_negative_y_array,
'v1': L_energy_array_ave,
'v2': loaded_data[prefix + 'FEDU_L' + suffix]['v2']},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_L' + suffix]['CDF']})
options(prefix + 'FEDU_L' + suffix, 'spec', 1)
# set ylim
ylim(prefix + 'FEDU_L' + suffix, 0, 180)
# set zlim
zlim(prefix + 'FEDU_L' + suffix, 1e+2, 1e+6)
tplot_variables.append(prefix + 'FEDU_L' + suffix)
# make Tplot Variables of erg_hep_l3_FEDU_L_paspec_ene?? (??: 00, 01, 02, ..., 15)
for i in range(loaded_data[prefix + 'FEDU_L' + suffix]['y'].shape[1]):
tplot_name = prefix + 'FEDU_L_paspec_ene' + \
str(i).zfill(2) + suffix
store_data(tplot_name, data={'x': loaded_data[prefix + 'FEDU_L' + suffix]['x'],
'y': non_negative_y_array[:, i, :],
'v': loaded_data[prefix + 'FEDU_L' + suffix]['v2']},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_L' + suffix]['CDF']})
options(tplot_name, 'spec', 1)
# set ylim
ylim(tplot_name, 0, 180)
# set zlim
zlim(tplot_name, 1e+2, 1e+6)
# set ytitle
options(
tplot_name, 'ytitle', f'HEP-L\nEne{str(i).zfill(2)}\n{L_energy_array[i]} keV')
tplot_variables.append(tplot_name)
if prefix + 'FEDU_H' + suffix in loaded_data:
H_energy_array_ave = np.sqrt(loaded_data[prefix + 'FEDU_H' + suffix]['v1'][0, :] *
loaded_data[prefix + 'FEDU_H' + suffix]['v1'][1, :]) # geometric mean for 'v1'
# get energy [keV] array for ytitle options
H_energy_array = np.trunc(H_energy_array_ave).astype(int)
non_negative_y_array = np.where(
loaded_data[prefix + 'FEDU_H' + suffix]['y'] < 0., np.nan, loaded_data[prefix + 'FEDU_H' + suffix]['y'])
store_data(prefix + 'FEDU_H' + suffix, data={'x': loaded_data[prefix + 'FEDU_H' + suffix]['x'],
'y': non_negative_y_array,
'v1': H_energy_array_ave,
'v2': loaded_data[prefix + 'FEDU_H' + suffix]['v2']},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_H' + suffix]['CDF']})
options(prefix + 'FEDU_H' + suffix, 'spec', 1)
# set ylim
ylim(prefix + 'FEDU_H' + suffix, 0, 180)
# set zlim
zlim(prefix + 'FEDU_H' + suffix, 1e+1, 1e+4)
tplot_variables.append(prefix + 'FEDU_H' + suffix)
# make Tplot Variables of erg_hep_l3_FEDU_H_paspec_ene?? (??: 00, 01, 02, ..., 10)
for i in range(loaded_data[prefix + 'FEDU_H' + suffix]['y'].shape[1]):
tplot_name = prefix + 'FEDU_H_paspec_ene' + \
str(i).zfill(2) + suffix
store_data(tplot_name, data={'x': loaded_data[prefix + 'FEDU_H' + suffix]['x'],
'y': non_negative_y_array[:, i, :],
'v': loaded_data[prefix + 'FEDU_H' + suffix]['v2']},
attr_dict={'CDF':loaded_data[prefix + 'FEDU_H' + suffix]['CDF']})
options(tplot_name, 'spec', 1)
# set ylim
ylim(tplot_name, 0, 180)
# set zlim
zlim(tplot_name, 1e+1, 1e+4)
# set ytitle
options(
tplot_name, 'ytitle', f'HEP-H\nEne{str(i).zfill(2)}\n{H_energy_array[i]} keV')
tplot_variables.append(tplot_name)
# set z axis to logscale
options(tplot_variables, 'zlog', 1)
# change colormap option
options(tplot_variables, 'colormap', 'jet')
# set ysubtitle
options(tplot_variables, 'ysubtitle', 'PA [deg]')
# set ztitle
options(tplot_variables, 'ztitle', '[/keV/cm^{2}/sr/s]')
return tplot_variables
return loaded_data
def dec2date(indata,
calendar='standard',
refdate=None,
units=None,
excelerr=True,
fulldate=None,
yr=False,
mo=False,
dy=False,
hr=False,
mi=False,
sc=False,
ascii=False,
en=False,
eng=False):
"""
Converts scale and array_like with decimal dates into
calendar dates. Supported time formats are:
standard, gregorian, julian, proleptic_gregorian,
excel1900, excel1904, 365_day, noleap, 366_day, all_leap,
and 360_day.
Input is decimal date in units of days.
Output is year, month, day, hour, minute, second
or any combination of them. Output in string format is possible.
Parameters
----------
indata : array_like
Input decimal dates. Dates must be positive.
calendar : str, optional
Calendar of input dates (default: 'standard').
Possible values are:
'standard', 'gregorian' = julian calendar from
01.01.-4712 12:00:00 (BC) until 05.03.1583 00:00:00 and
gregorian calendar from 15.03.1583 00:00:00 until now.
Missing 10 days do not exsist.
'julian' = julian calendar from 01.01.-4712 12:00:00 (BC)
until now.
'proleptic_gregorian' = gregorian calendar from
01.01.0001 00:00:00 until now.
'excel1900' = Excel dates with origin at
01.01.1900 00:00:00.
'excel1904' = Excel 1904 (Lotus) format.
Same as excel1904 but with origin at
01.01.1904 00:00:00.
'365_day', 'noleap' = 365 days format,
i.e. common years only (no leap years)
with origin at 01.01.0001 00:00:00.
'366_day', 'all_leap' = 366 days format,
i.e. leap years only (no common years)
with origin at 01.01.0001 00:00:00.
'360_day' = 360 days format,
i.e. years with only 360 days (30 days per month)
with origin at 01.01.0001 00:00:00.
'decimal' = decimal year instead of decimal days.
'decimal360' = decimal year with a year of 360 days, i.e. 12 month with 30 days each.
Optional Arguments
------------------
refdate : str, optional
Reference date for 'days since refdate' can be set by user. Input must be a
string in the format 'yyyy-mm-dd hh:mm:ss'.
Default values for different `calendars` are set automatically.
units : str, optional
Units of the the time stamp can be given. This can only be the following two:
'day as %Y%m%d.%f', or
'days since refdate', with refdate in the format 'yyyy-mm-dd hh:mm:ss'.
If `units='day as %Y%m%d.%f'` then `calendar` is ignored and the dates will be returned
assuming that %f is fractional date from 00:00:00 h.
excelerr : bool, optional
In Excel, the year 1900 is normally considered a leap year,
which it was not. By default, this error is taken into account
if calendar='excel1900' (default: True).
1900 is not considered a leap year if excelerr=False.
Returns
-------
list of array_like
`fulldate` -> output arrays with year, month, day, hour, minute, second
Default fulldate is overwritten by selection of special output `yr`,`mn`,`dy`,`hr`,`mi`,`sc`
`yr` -> output array with year
`mo` -> output array with month
`dy` -> output array with day
`hr` -> output array with hour
`mi` -> output array with minute
`sc` -> output array with second
`ascii` -> output array with strings of the format 'dd.mm.yyyy hh:mm:ss'
`en` -> output array with strings of the format 'yyyy-mm-dd hh:mm:ss'
`eng` -> Same as en: obsolete.
Notes
-----
Most versions of `datetime do` not support negative years,
i.e. Julian days < 1721423.5 = 01.01.0001 00:00.
There is an issue in `netcdftime` version < 0.9.5 in proleptic_gregorian for dates before year 301:
dec2date(date2dec(ascii='01.01.0300 00:00:00', calendar='proleptic_gregorian'), calendar='proleptic_gregorian')
[300, 1, 2, 0, 0, 0]
dec2date(date2dec(ascii='01.01.0301 00:00:00', calendar='proleptic_gregorian'), calendar='proleptic_gregorian')
[301, 1, 1, 0, 0, 0]
Requires `netcdftime.py` from module netcdftime available at:
http://netcdf4-python.googlecode.com
Examples
--------
# Some implementations of datetime have problems with negative years
>>> import datetime
>>> if datetime.MINYEAR > 0:
... print('The minimum year in your datetime implementation is ', datetime.MINYEAR)
... print('i.e. it does not support negative years (BC).')
The minimum year in your datetime implementation is 1
i.e. it does not support negative years (BC).
#calendar = 'standard'
>>> year = np.array([2000,1810,1630,1510,1271,619,1])
>>> month = np.array([1,4,7,9,3,8,1])
>>> day = np.array([5,24,15,20,18,27,1])
>>> hour = np.array([12,16,10,14,19,11,12])
>>> minute = np.array([30,15,20,35,41,8,0])
>>> second = np.array([15,10,40,50,34,37,0])
>>> from date2dec import date2dec
>>> decimal = date2dec(calendar='standard', yr=year, mo=month, dy=day, hr=hour, mi=minute, sc=second)
>>> year1, month1, day1, hour1, minute1, second1 = dec2date(decimal, calendar= 'standard', fulldate=True)
>>> print(year1)
[2000 1810 1630 1510 1271 619 1]
>>> print(month1)
[1 4 7 9 3 8 1]
>>> print(day1)
[ 5 24 15 20 18 27 1]
>>> print(hour1)
[12 16 10 14 19 11 12]
>>> print(minute1)
[30 15 20 35 41 8 0]
>>> print(second1)
[15 10 40 50 34 37 0]
# calendar = 'julian'
>>> decimal = date2dec(calendar='julian', yr=year, mo=month, dy=day, hr=hour, mi=minute, sc=second)
>>> year1 = dec2date(decimal, calendar='julian', yr=True)
>>> print(year1)
[2000 1810 1630 1510 1271 619 1]
# calendar = 'proleptic_gregorian'
>>> decimal = date2dec(calendar='proleptic_gregorian', yr=year, mo=month, dy=day, hr=hour, mi=minute, sc=second)
>>> ascii = dec2date(decimal, calendar='proleptic_gregorian', ascii=True)
>>> print(ascii[::4])
['05.01.2000 12:30:15' '18.03.1271 19:41:34']
# calendar = 'excel1900' WITH excelerr = True -> 1900 considered as leap year
>>> decimal = date2dec(calendar='excel1900', yr=year, mo=month, dy=day, hr=hour, mi=minute, sc=second)
>>> year1, day1 = dec2date(decimal, calendar='excel1900', yr=True, dy=True)
>>> print(year1)
[2000 1810 1630 1510 1271 619 1]
>>> print(day1)
[ 5 24 15 20 18 27 1]
# calendar = 'excel1900' WITH excelerr = False -> 1900 considered as NO leap year
# Older versions of netcdftime.py produced unnecessary output (Line 262)
# >>> decimal = date2dec(calendar='excel1900',yr=year,mo=month,dy=day,hr=hour,mi=minute,sc=second,excelerr=False)
# >>> if nt.__version__ < '0.9.4':
# ... asciidate = dec2date(decimal, calendar='excel1900', ascii = True, excelerr = False)
# ... elif nt.__version__ == '0.9.4':
# ... asciidate = dec2date(decimal, calendar='excel1900', ascii = True, excelerr = False)
# ... for i in range(3):
# ... print('0 300')
# ... else:
# ... asciidate = dec2date(decimal, calendar='excel1900', ascii = True, excelerr = False)
# ... for i in range(7):
# ... print('0 300')
# 0 300
# 0 300
# 0 300
# 0 300
# 0 300
# 0 300
# 0 300
# >>> print(asciidate[::4])
# ['05.01.2000 12:30:15' '18.03.1271 19:41:34']
# calendar = 'excel1904'
>>> decimal = date2dec(calendar='excel1904', yr=year, mo=month, dy=day, hr=hour, mi=minute, sc=second)
>>> asciidate = dec2date(decimal, calendar='excel1904', ascii = True)
>>> print(asciidate[::4])
['05.01.2000 12:30:15' '18.03.1271 19:41:34']
>>> asciidate = dec2date(decimal, calendar='excel1904', ascii=True, refdate='1909-12-31 00:00:00')
>>> print(asciidate[::4])
['05.01.2006 12:30:15' '18.03.1277 19:41:34']
>>> print(dec2date(decimal[::4], calendar='excel1904', ascii=True, units='days since 1909-12-31 00:00:00'))
['05.01.2006 12:30:15' '18.03.1277 19:41:34']
>>> print(dec2date(decimal[::4], calendar='excel1904', ascii=True, units='days since 1910-01-01 00:00:00'))
['06.01.2006 12:30:15' '19.03.1277 19:41:34']
# check especially 1900 (no) leap year in Excel
>>> year1 = np.array([1900,1900,1900,1900])
>>> month1 = np.array([2,2,3,1])
>>> day1 = np.array([28,29,1,1])
>>> decimal = date2dec(calendar='excel1900', yr=year1, mo=month1, dy=day1)
>>> month2, day2 = dec2date(decimal, calendar='excel1900', mo=True, dy=True)
>>> print(month2)
[2 2 3 1]
>>> print(day2)
[28 29 1 1]
>>> decimal = date2dec(calendar='excel1900', yr=year1, mo=month1, dy=day1, excelerr=False)
>>> month2, day2 = dec2date(decimal, calendar='excel1900', mo=True, dy=True, excelerr=False)
>>> print(month2)
[2 3 3 1]
>>> print(day2)
[28 1 1 1]
>>> decimal = date2dec(calendar='excel1904', yr=year1, mo=month1, dy=day1)
>>> month2, day2 = dec2date(decimal, calendar='excel1904', mo=True, dy=True)
>>> print(month2)
[2 3 3 1]
>>> print(day2)
[28 1 1 1]
# calendar = '365_day'
>>> decimal = date2dec(calendar='365_day',yr=year,mo=month,dy=day,hr=hour,mi=minute,sc=second)
>>> asciidate = dec2date(decimal, calendar='365_day', ascii = True)
>>> print(asciidate[::4])
['05.01.2000 12:30:15' '18.03.1271 19:41:34']
# calendar = '366_day'
>>> decimal = date2dec(calendar='366_day',yr=year,mo=month,dy=day,hr=hour,mi=minute,sc=second)
>>> asciidate = dec2date(decimal, calendar='366_day', ascii = True)
>>> print(asciidate[::4])
['05.01.2000 12:30:15' '18.03.1271 19:41:34']
# calendar = '360_day'
>>> decimal = date2dec(calendar='360_day',yr=year,mo=month,dy=day,hr=hour,mi=minute,sc=second)
>>> asciidate = dec2date(decimal, calendar='360_day', ascii = True)
>>> print(asciidate[::4])
['05.01.2000 12:30:15' '18.03.1271 19:41:34']
>>> print(dec2date(719644.52101, calendar='proleptic_gregorian', ascii = True))
28.04.1971 12:30:15
>>> dec = date2dec(ascii='02.03.1910 03:44:55', calendar='decimal')
>>> print(dec2date(dec, calendar='decimal', ascii=True))
02.03.1910 03:44:55
>>> dec = date2dec(ascii='02.03.1910 03:44:55', calendar='decimal360')
>>> print(dec2date(dec, calendar='decimal360', ascii=True))
02.03.1910 03:44:55
>>> print(dec2date([dec,dec], calendar='decimal360', ascii=True))
['02.03.1910 03:44:55', '02.03.1910 03:44:55']
>>> print(dec2date([[dec,dec],[dec,dec],[dec,dec]], calendar='decimal360', ascii=True)[0])
['02.03.1910 03:44:55', '02.03.1910 03:44:55']
>>> print(dec2date(np.array([dec,dec]), calendar='decimal360', ascii=True))
['02.03.1910 03:44:55' '02.03.1910 03:44:55']
>>> print(dec2date(np.array([[dec,dec],[dec,dec],[dec,dec]]), calendar='decimal360', ascii=True)[0:2,0])
['02.03.1910 03:44:55' '02.03.1910 03:44:55']
>>> absolut = np.array([20070102.0034722, 20070102.0069444])
>>> print(dec2date(absolut, units='day as %Y%m%d.%f', ascii=True))
['02.01.2007 00:05:00' '02.01.2007 00:10:00']
>>> absolut = [20070102.0034722, 20070102.0069444]
>>> print(dec2date(absolut, units='day as %Y%m%d.%f', ascii=True))
['02.01.2007 00:05:00', '02.01.2007 00:10:00']
>>> absolut = np.array([200401.5, 200402.5, 201011.5, 201002.5])
>>> print(dec2date(absolut, units='month as %Y%m.%f', ascii=True))
['15.01.2004 12:00:00' '14.02.2004 12:00:00' '15.11.2010 00:00:00' '14.02.2010 00:00:00']
>>> absolut = np.array([2004.5, 2010.5])
>>> print(dec2date(absolut, units='year as %Y.%f', ascii=True))
['01.07.2004 00:00:00' '01.07.2010 12:00:00']
# en, eng
>>> print(dec2date(719644.52101, calendar='proleptic_gregorian', en=True))
1971-04-28 12:30:15
>>> print(dec2date(719644.52101, calendar='proleptic_gregorian', eng=True))
1971-04-28 12:30:15
History
-------
Written Arndt Piayda, Jun 2010
Modified Matthias Cuntz, Feb 2012 - Input can be scalar or array
- Default: fulldate=True
- Changed checks for easier extension
- decimal, decimal360
Matthias Cuntz, Jun 2012 - former units keyword is now called refdate
- units has now original meaning as in netcdftime
- units='day as %Y%m%d.%f'
Matthias Cuntz, Feb 2013 - solved Excel leap year problem.
Matthias Cuntz, Feb 2013 - ported to Python 3
Arndt Piayda, May 2013 - solved eng output problem.
Matthias Cuntz, Oct 2013 - Excel starts at 1 not at 0
Matthias Cuntz, Oct 2013 - units bugs, e.g. 01.01.0001 was substracted if Julian calendar even with units
Matthias Cuntz, May 2016 - units=='month as %Y%m.%f', units=='year as %Y.%f'
Matthias Cuntz, Oct 2016 - netcdftime provided even with netCDF4 > 1.0.0; make leap always integer
Matthias Cuntz, May 2020 - numpy docstring format
Matthias Cuntz, Jul 2020 - en for eng
Matthias Cuntz, Jul 2020 - use proleptic_gregorian for Excel dates
"""
#
# Constants
calendars = [
'standard', 'gregorian', 'julian', 'proleptic_gregorian', 'excel1900',
'excel1904', '365_day', 'noleap', '366_day', 'all_leap', '360_day',
'decimal', 'decimal360'
]
#
# Checks
import netCDF4 as nt
try:
tst = nt.date2num
tst = nt.datetime
except:
try:
import netcdftime as nt
if ((nt.__version__ <= '0.9.2') & (calendar == '360_day')):
raise ValueError(
"date2dec error: Your version of netcdftime.py is equal"
" or below 0.9.2. The 360_day calendar does not work with"
" arrays here. Please download a newer one.")
except:
import cftime as nt
#
calendar = calendar.lower()
if (calendar not in calendars):
raise ValueError("dec2date error: Wrong calendar! Choose: " +
''.join([i + ' ' for i in calendars]))
if refdate and units:
raise ValueError(
"dec2date error: either refdate or units can be given.")
# obsolete eng
if en and eng:
raise ValueError(
"dec2date error: 'eng' was succeeded by 'en'. Only one can be given."
)
if eng and (not en): en = eng
#
# Default
if np.sum(np.array([yr, mo, dy, hr, mi, sc])) >= 1:
ii = True
else:
ii = False
if ((ascii | en | ii) and (not fulldate)):
fulldate = False
if ((not (ascii | en | ii)) and (not fulldate)):
fulldate = True
if fulldate == True:
yr = True
mo = True
dy = True
hr = True
mi = True
sc = True
# Further checks
if np.sum(np.array([ascii, fulldate, en])) > 1:
raise ValueError(
"dec2date error: Only one of ascii, fulldate, or en can be chosen."
)
if np.sum(np.array([ascii, en, ii])) > 1:
raise ValueError(
"dec2date error: If ascii, fulldate or en then no special selection yr,mo,dy,hr,mi,sc possible."
)
#
# Input size and shape
islist = type(indata) != type(np.array(indata))
isarr = np.ndim(indata)
if (islist & (isarr > 2)):
raise ValueError("dec2date error: input is list > 2D; Use array input")
if isarr == 0: indata = np.array([indata])
else: indata = np.array(indata)
insize = indata.size
inshape = indata.shape
indata = indata.flatten()
#
# depending on chosen calendar and optional set of the time refdate
# calendar date is calculated
if units == 'day as %Y%m%d.%f':
fdy = indata % 1. # day fraction
indata = indata - fdy
day = np.rint(indata % 100.).astype(np.int)
indata = indata - day
tmp = indata % 10000.
month = np.rint(tmp / 100.).astype(np.int)
indata = indata - tmp
year = np.rint(indata / 10000.).astype(np.int)
secs = np.rint(fdy * 86400.)
hour = np.floor(secs / 3600.).astype(np.int)
secs = secs - 3600. * hour
minute = np.floor(secs / 60.).astype(np.int)
second = np.rint(secs - 60. * minute).astype(np.int)
elif units == 'month as %Y%m.%f':
fmo = indata % 1. # month fraction
indata = indata - fmo
month = np.rint(indata % 100.).astype(np.int)
indata = indata - month
year = np.rint(indata / 100.).astype(np.int)
leap = np.where(
(((year % 4) == 0) & ((year % 100) != 0)) | ((year % 400) == 0), 1,
0)
dim = np.array([[-9, 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31],
[-9, 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31]])
indata = dim[(leap, month)] * fmo
fdy = indata % 1. # day fraction
indata = indata - fdy
day = np.rint(indata % 100.).astype(np.int)
secs = np.rint(fdy * 86400.)
hour = np.floor(secs / 3600.).astype(np.int)
secs = secs - 3600. * hour
minute = np.floor(secs / 60.).astype(np.int)
second = np.rint(secs - 60. * minute).astype(np.int)
elif units == 'year as %Y.%f':
fyr = indata % 1. # year fraction
year = np.rint(indata - fyr).astype(np.int)
leap = np.where(
(((year % 4) == 0) & ((year % 100) != 0)) | ((year % 400) == 0), 1,
0)
dsiy = np.array([365, 366])
indata = dsiy[leap] * fyr
fdy = indata % 1. # day fraction
doy = np.rint(indata - fdy).astype(np.int)
diy = np.array(
[[-9, 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365],
[-9, 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366]])
month = np.zeros(insize, dtype=np.int)
day = np.zeros(insize, dtype=np.int)
for i in range(insize):
month[i] = np.where(doy[i] > np.squeeze(diy[leap[i], :]))[0][-1]
day[i] = doy[i] - diy[leap[i], month[i]]
secs = np.rint(fdy * 86400.)
hour = np.floor(secs / 3600.).astype(np.int)
secs = secs - 3600. * hour
minute = np.floor(secs / 60.).astype(np.int)
second = np.rint(secs - 60. * minute).astype(np.int)
else:
if (calendar == 'standard') or (calendar == 'gregorian'):
dec0 = 0
if units:
unit = units
elif refdate:
#unit = 'days since %s' % (refdate)
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 12:00:00'
dec0 = 1721424
timeobj = nt.num2date(indata - dec0, unit, calendar='gregorian')
elif calendar == 'julian':
dec0 = 0
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 12:00:00'
dec0 = 1721424
timeobj = nt.num2date(indata - dec0, unit, calendar='julian')
elif calendar == 'proleptic_gregorian':
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 00:00:00'
timeobj = nt.num2date(indata, unit, calendar='proleptic_gregorian')
elif calendar == 'excel1900':
doerr = False
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 1899-12-31 00:00:00'
if excelerr: doerr = True
if doerr:
indata1 = np.where(indata >= 61., indata - 1, indata)
timeobj = nt.num2date(indata1,
unit,
calendar='proleptic_gregorian')
else:
timeobj = nt.num2date(indata,
unit,
calendar='proleptic_gregorian')
elif calendar == 'excel1904':
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 1903-12-31 00:00:00'
timeobj = nt.num2date(indata, unit, calendar='proleptic_gregorian')
elif (calendar == '365_day') or (calendar == 'noleap'):
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 00:00:00'
timeobj = nt.num2date(indata, unit, calendar='365_day')
elif (calendar == '366_day') or (calendar == 'all_leap'):
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 00:00:00'
timeobj = nt.num2date(indata, unit, calendar='366_day')
elif calendar == '360_day':
if units:
unit = units
elif refdate:
unit = 'days since {0:s}'.format(refdate)
else:
unit = 'days since 0001-01-01 00:00:00'
timeobj = nt.num2date(indata, unit, calendar='360_day')
elif calendar == 'decimal':
fyear = np.trunc(indata)
year = np.array(fyear, dtype=np.int)
leap = ((((year % 4) == 0) & ((year % 100) != 0)) |
((year % 400) == 0)).astype(np.int)
fleap = leap.astype(np.float)
fract_date = indata - fyear
days_year = 365.
# date in hours
fhoy = fract_date * (days_year + fleap) * 24.
fihoy = np.trunc(fhoy)
ihoy = np.array(fihoy, dtype=np.int)
# minutes
fmoy = (fhoy - fihoy) * 60.
fminute = np.trunc(fmoy)
minute = np.array(fminute, dtype=np.int)
# seconds
second = np.array(np.trunc((fmoy - fminute) * 60.), dtype=np.int)
# months
fdoy = (fihoy / 24.) + 1.
fidoy = np.trunc(fdoy)
idoy = np.array(fidoy, dtype=np.int)
month = np.zeros(insize, dtype=np.int)
diy = np.array([[
-9, 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365
], [
-9, 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366
]])
for i in range(insize):
ii = np.squeeze(
np.where(idoy[i] > np.squeeze(diy[leap[i], :])))
month[i] = ii[-1]
# days
fday = np.zeros(insize, dtype=np.float)
for i in range(insize):
fday[i] = np.trunc(fdoy[i] - np.float(diy[leap[i], month[i]]))
day = np.array(fday, dtype=np.int)
# hours
hour = ihoy % 24
elif calendar == 'decimal360':
fyear = np.trunc(indata)
year = np.array(fyear, dtype=np.int)
fract_date = indata - fyear
days_year = 360.
# date in hours
fhoy = fract_date * days_year * 24.
fihoy = np.trunc(fhoy)
ihoy = np.array(fihoy, dtype=np.int)
# minutes
fmoy = (fhoy - fihoy) * 60.
fminute = np.trunc(fmoy)
minute = np.array(fminute, dtype=np.int)
# seconds
second = np.array(np.trunc((fmoy - fminute) * 60.), dtype=np.int)
# months
fdoy = (fihoy / 24.) + 1.
fidoy = np.trunc(fdoy)
idoy = np.array(fidoy, dtype=np.int)
month = np.zeros(insize, dtype=np.int)
diy = np.array([
-9, 0, 30, 60, 90, 120, 150, 180, 210, 240, 270, 300, 330, 360
])
for i in range(insize):
ii = np.squeeze(np.where(idoy[i] > diy))
month[i] = ii[-1]
# days
fday = np.zeros(insize, dtype=np.float)
for i in range(insize):
fday[i] = np.trunc(fdoy[i] - np.float(diy[month[i]]))
day = np.array(fday, dtype=np.int)
# hours
hour = ihoy % 24
else:
raise ValueError(
"dec2date error: calendar not implemented; should have been catched before."
)
if (calendar not in ['decimal', 'decimal360']):
timeobjfl = timeobj.flatten()
year = np.array([timeobjfl[i].year for i in range(insize)],
dtype=np.int)
month = np.array([timeobjfl[i].month for i in range(insize)],
dtype=np.int)
day = np.array([timeobjfl[i].day for i in range(insize)],
dtype=np.int)
hour = np.array([timeobjfl[i].hour for i in range(insize)],
dtype=np.int)
minute = np.array([timeobjfl[i].minute for i in range(insize)],
dtype=np.int)
second = np.array([timeobjfl[i].second for i in range(insize)],
dtype=np.int)
if (calendar == 'excel1900') & excelerr:
ii = np.where((indata >= 60.) & (indata < 61.))[0]
if np.size(ii) > 0:
month[ii] = 2
day[ii] = 29
#
# Ascii output
if ascii:
output = ([
'%02d.%02d.%04d %02d:%02d:%02d' %
(day[i], month[i], year[i], hour[i], minute[i], second[i])
for i in range(insize)
])
output = np.reshape(output, inshape)
if isarr == 0:
output = output[0]
# Ascii english output
elif en:
output = ([
'%04d-%02d-%02d %02d:%02d:%02d' %
(year[i], month[i], day[i], hour[i], minute[i], second[i])
for i in range(insize)
])
output = np.reshape(output, inshape)
if isarr == 0:
output = output[0]
else:
# Individual output
# if one, some or all of yr, mo, dy, hr, mi or sc is
# choosen by the user as output, arrays for datetime
year = np.reshape(year, inshape)
month = np.reshape(month, inshape)
day = np.reshape(day, inshape)
hour = np.reshape(hour, inshape)
minute = np.reshape(minute, inshape)
second = np.reshape(second, inshape)
if isarr == 0:
year = np.int(year)
month = np.int(month)
day = np.int(day)
hour = np.int(hour)
minute = np.int(minute)
second = np.int(second)
# filling of output list:
output = []
if yr: output += [year]
if mo: output += [month]
if dy: output += [day]
if hr: output += [hour]
if mi: output += [minute]
if sc: output += [second]
# return output arrays:
if len(output) == 1:
output = output[0]
if isarr != 0:
if islist:
ns = np.size(inshape)
if ns == 1:
output = [i for i in output]
elif ns == 2:
loutput = [i for i in output[:, 0]]
for i in range(np.size(output[:, 0])):
loutput[i] = list(np.squeeze(output[i, :]))
output = loutput
else:
raise ValueError(
"dec2date error: list output > 2D; should have been catched before."
)
return output
def scalar_trunc(x: Number) -> Number:
"""Implement `scalar_trunc`."""
_assert_scalar(x)
return np.trunc(x)
def __init__(self, dim, N=4.5):
# Log spaced CR from 1/dim to dim/dim
self._set_CRs(logspace(0, log10(dim), trunc(N * log10(dim) + 1)) / dim)
def slide_window_search(binary_warped, left_current, right_current):
out_img = np.dstack((binary_warped, binary_warped, binary_warped))
nwindows = 4
window_height = np.int(binary_warped.shape[0] / nwindows)
nonzero = binary_warped.nonzero()
nonzero_y = np.array(nonzero[0])
nonzero_x = np.array(nonzero[1])
margin = 100
minpix = 50
left_lane = []
right_lane = []
color = [0, 255, 0]
thickness = 2
for w in range(nwindows):
win_y_low = binary_warped.shape[0] - (w + 1) * window_height
win_y_high = binary_warped.shape[0] - w * window_height
win_xleft_low = left_current - margin
win_xleft_high = left_current + margin
win_xright_low = right_current - margin
win_xright_high = right_current + margin
cv2.rectangle(out_img, (win_xleft_low, win_y_low),
(win_xleft_high, win_y_high), color, thickness)
cv2.rectangle(out_img, (win_xright_low, win_y_low),
(win_xright_high, win_y_high), color, thickness)
good_left = ((nonzero_y >= win_y_low) & (nonzero_y < win_y_high) &
(nonzero_x >= win_xleft_low) &
(nonzero_x < win_xleft_high)).nonzero()[0]
good_right = ((nonzero_y >= win_y_low) & (nonzero_y < win_y_high) &
(nonzero_x >= win_xright_low) &
(nonzero_x < win_xright_high)).nonzero()[0]
left_lane.append(good_left)
right_lane.append(good_right)
if len(good_left) > minpix:
left_current = np.int(np.mean(nonzero_x[good_left]))
if len(good_right) > minpix:
right_current = np.int(np.mean(nonzero_x[good_right]))
left_lane = np.concatenate(left_lane)
right_lane = np.concatenate(right_lane)
leftx = nonzero_x[left_lane]
lefty = nonzero_y[left_lane]
rightx = nonzero_x[right_lane]
righty = nonzero_y[right_lane]
left_fit = np.polyfit(lefty, leftx, 2)
right_fit = np.polyfit(righty, rightx, 2)
ploty = np.linspace(0, binary_warped.shape[0] - 1, binary_warped.shape[0])
left_fitx = left_fit[0] * ploty**2 + left_fit[1] * ploty + left_fit[2]
right_fitx = right_fit[0] * ploty**2 + right_fit[1] * ploty + right_fit[2]
ltx = np.trunc(left_fitx)
rtx = np.trunc(right_fitx)
out_img[nonzero_y[left_lane], nonzero_x[left_lane]] = [255, 0, 0]
out_img[nonzero_y[right_lane], nonzero_x[right_lane]] = [0, 0, 255]
if len(leftx) < len(rightx):
mid_size = len(leftx)
midx = (rightx[:mid_size] + leftx) // 2
midy = (righty[:mid_size] + lefty) // 2
else:
mid_size = len(rightx)
midx = (rightx[:mid_size] + rightx) // 2
midy = (righty[:mid_size] + righty) // 2
new_midx = int(np.mean(midx) / 2)
new_midy = int(np.mean(midy) / 2)
new_mid = (new_midx, new_midy)
# print(new_mid, frame_size)
ret = {
'left_fitx': ltx,
'right_fitx': rtx,
'ploty': ploty,
'new_mid': new_mid
}
return ret
ax3.axhline(y=0, color='k')
ax3.legend(("Home", "Away"), loc=0)
ax4 = plt.subplot(2, 3, 4)
ax4.set_title("Average Wins v. Goals in MLS")
ax4.set_xlabel("# of Goals")
ax4.set_ylabel("# of Wins")
ax4.set_xticks(range(0, np.max(hGoals) + 1, 30))
ax4.set_yticks(range(0, np.max(hWins), 10))
ax4.plot(hGoals, hWins, 'o', color='blue')
ax4.plot(vGoals, vWins, 'o', color='red')
ax4.plot(hGoals, modelH.fittedvalues, 'b--', label='Home OLS')
ax4.plot(vGoals, modelV.fittedvalues, 'r--', label='Vis OLS')
ax4.set_ylim(ymin=0)
ax4.axhline(y=0, color='k')
ax4.text(360, 40, ("Home R^2 = {}%".format(np.trunc(100 * modelH.rsquared))))
ax4.text(360, 20, ("Vis R^2 = {}%".format(np.trunc(100 * modelV.rsquared))))
ax4.legend(("Home", "Away"), loc=0)
ax5 = plt.subplot(2, 3, 5)
ax5.set_title("Home v. Away Wins in MLS")
ax5.set_xlabel("Year")
ax5.set_ylabel("# of Wins")
ax5.set_xticks(range(np.min(years), np.max(years) + 1, 2))
ax5.set_yticks(range(0, 300, 50))
ax5.plot(years, hWins, color='blue', linestyle='solid')
ax5.plot(years, vWins, color='red', linestyle='solid')
ax5.set_ylim(ymin=0)
ax5.axhline(y=0, color='k')
ax5.legend(("Home", "Away"), loc=2)
#but home and away goals aren't the whole story. What about goals per # of teams? More teams each year so see if goals per team increases
fh.write('# Pixel values in e-/sec\n')
fh.write(' '.join(colnames) + '\n')
for time in np.arange(now.secs, now.secs + 3600, 4.1):
print(time - now.secs)
t_ccd = get_t_ccd(time)
scale = dark_cal.dark_temp_scale(t_ccd0, t_ccd_ref=t_ccd)
pix_readout_electrons = pixels * scale * t_readout
# Add gaussian count noise
count_noise = np.sqrt(pix_readout_electrons)
pix_readout_electrons += np.random.normal(loc=0,
scale=count_noise,
size=128)
pix_readout_dn = np.trunc(pix_readout_electrons / 5)
first_image = pix_readout_dn.tolist()[0:64]
second_image = pix_readout_dn.tolist()[64:]
for image, pix_filename, slot in zip(
[first_image, second_image], [opt.pix_filename1, opt.pix_filename2],
[6, 7]):
vals = [time, t_ccd, slot, t_readout] + image
vals = ['{:.2f}'.format(val) for val in vals]
with open(pix_filename, 'a') as fh:
fh.write(' '.join(vals) + '\n')
if opt.delay:
sleep(opt.delay)
print("yesterday: ", yesterday)
print("today: ", today)
print("tomorrow: ", tomorrow)
# #### 58. 使用五种不同的方法去提取一个随机数组的整数部分
# In[67]:
Z = np.random.uniform(0, 10, 10)
print("原始值: ", Z)
print("方法 1: ", Z - Z % 1)
print("方法 2: ", np.floor(Z))
print("方法 3: ", np.ceil(Z) - 1)
print("方法 4: ", Z.astype(int))
print("方法 5: ", np.trunc(Z))
# In[85]:
Z = np.random.uniform(0, 10, 10)
print("origin:", Z)
print("1: ", Z - Z % 1)
print("2: ", np.floor(Z))
print("3: ", np.ceil(Z) - 1)
print("4: ", Z.astype(int))
print("5: ", np.trunc(Z))
# #### 59. 创建一个 5x5 的矩阵，其中每行的数值范围从 1 到 5
# In[68]:
# cbox = (int((ymin+ymax)/2),int((xmin+xmax)/2)) #centro:(Y,X)
# scentros.append(cbox)
# dbox = (ymax-ymin,xmax-xmin)
# sdims.append(dbox) #dimension: (alto,ancho)
# #print('Silla ',index,' [Xmin,Ymin,Xmax,Ymax]=',rbox,' Centro=',cbox,' Dimensiones:',dbox)
s = [[163, 279, 441, 484], [143, 292, 289, 482], [87, 201, 295, 338],
[108, 323, 276, 500], [107, 198, 252, 305]]
c = [(302, 381), (216, 387), (191, 269), (192, 411), (179, 251)]
d = [(278, 205), (146, 190), (208, 137), (168, 177), (145, 107)]
dmin = [[145, 107]]
#print(s,c,d)
sillas = np.array(s)
scentros = np.trunc(
np.concatenate(
(np.expand_dims(np.divide(sillas[:, 0] + sillas[:, 2], 2), axis=1),
np.expand_dims(np.divide(sillas[:, 1] + sillas[:, 3], 2), axis=1)),
axis=1))
sdims = np.concatenate((np.expand_dims(sillas[:, 2] - sillas[:, 0], axis=1),
np.expand_dims(sillas[:, 3] - sillas[:, 1], axis=1)),
axis=1).astype('int32')
#print(sillas,scentros,sdims)
x = [[1., 0.39381477, 0.10023424, 0.26546335, 0.03296719],
[0.39381477, 1., 0.13562197, 0.58209695, 0.03386873],
[0.10023424, 0.13562197, 1., 0.04523262, 0.52124019],
[0.26546335, 0.58209695, 0.04523262, 1., 0.],
[0.03296719, 0.03386873, 0.52124019, 0., 1.]]
id = [[2., 3.], [1., 3.], [1., 2.], [1., 3.], [1., 2.]]
if True:
if len(sillas) > 0:
print('Sillas:\n', sillas)
def avgstrainfit(V,
E,
V0,
nmax=16,
MODE=1,
strain='eulerian',
LOG=0,
nargout=1):
"""avgstrainfit - Fit to an average strain polynomials."""
import sys
if (len(V) != len(E)):
sys.exit('avgstrainfit: V and E must be vectors of the same length!')
elif (len(V) < 7):
sys.exit('avgstrainfit: dataset must have at least 7 points!')
elif MODE != 1 and MODE != 2:
sys.exit('avgstrainfit: weighting mode must be 1 or 2!')
ndata = len(V)
Vrange = [np.amin(V), np.amax(V)]
# Determine the maximum degree of the polynomials:
if (nmax < 0):
MaxDegree = int(np.amin([ndata - 5, np.trunc(ndata / 2.)]))
else:
MaxDegree = int(
np.amin([nmax, np.amin([ndata - 5, np.trunc(ndata / 2.)])]))
# Some statistics of the averaging proccess:
Morder = 0
npol = 0
pol = []
s2 = []
for n in range(2, MaxDegree + 1):
c = strainfit(V, E, V0, n, strain, 0, nargout=1)
sm = strainmin(c, V0, Vrange, strain)
Efit = strainevalE(c, V0, V, strain)
SSerr = np.sum((E - Efit)**2)
pol.append(polynomial())
pol[npol].c = c
pol[npol].SSerr = SSerr
s2.append(SSerr)
pol[npol].order = n
pol[npol].data = ndata
pol[npol].smin = sm
Morder = np.max([n, Morder])
npol += 1
#print(c)
#print('n,Morder,npol = ',n,Morder,npol)
#print(sm)
#print(Efit)
#print(SSerr)
#print('s2 = ',s2)
#print(pol)
# Get the polynomial weights:
SSmin = np.amin(s2)
Q = 0
if (MODE == 1):
for k in range(npol):
w = (pol[k].SSerr / SSmin) * (pol[k].order / pol[k].data)
pol[k].w = np.exp(-w)
Q += pol[k].w
elif (MODE == 2):
for k in range(npol):
w = (pol[k].SSerr / SSmin) * (pol[k].order / pol[k].data)
pol[k].w = np.exp(-w * w)
Q += pol[k].w
##ww = zeros(1,1:npol);
ww = np.zeros(npol)
for k in range(npol):
ww[k] = pol[k].w / Q
pol[k].w = pol[k].w / Q
#print(pol)
# Form the average polynomial:
cavg = np.zeros(Morder + 1)
for k in range(npol):
n = pol[k].order
cavg[Morder - n:] += pol[k].c * pol[k].w
#"""
# Extra output and optional LOG record:
# If the average of polynomials has a minimum, analyze the equilibrium
# geometry.
# Otherwise, analyze the reference volume.
if (LOG > 0 or nargout > 1):
avgmin = strainmin(cavg, V0, Vrange, strain)
if (avgmin.err == 0):
# Analyze and report the equilibrium geometry:
if (LOG > 0):
print(
'\n\navgpolyfit: AVERAGE OF {0} STRAIN POLYNOMIALS'.format(
strain))
print('Volume reference (V0): {0:.6f}'.format(V0))
print('Range of degrees: 2--{0}'.format(MaxDegree))
print('Number of polynomials: {0}'.format(npol))
print('\nProperties at the minimum of each polynomial:')
print(
'--i-- npol data order --SSerr-- ---w---- ----Vmin--- -----Emin--- ----Bmin--- ---B1min--- ---B2min--- ---B3min---'
)
isrt = np.argsort(ww)[::-1]
srt = ww[isrt]
pmean = np.zeros(6)
pstd = np.zeros(6)
for k in range(npol):
k1 = isrt[k]
###[smin] = strainmin(pol{k1}.c, V0, Vrange, strain);
smin = pol[k1].smin
prop = straineval(pol[k1].c, V0, smin.Vmin, strain)
if (LOG > 0 and k <= 25):
# In the conversion of the bulk modulus and derivatives we
# assume the units: volume (bohr^3), energy (Hy).
output_string = "{0:4} {1:4} {2:4} {3:4}".format(
k, k1 + 1, pol[k1].data, pol[k1].order)
output_string += " {0:9.2e} {1:8.6f}".format(
pol[k1].SSerr, pol[k1].w)
output_string += " {0:10.6f}".format(smin.Vmin)
output_string += " {0:13.6f}".format(prop.E)
output_string += " {0:10.6f}".format(prop.B *
hybohr3togpa)
output_string += " {0:9.6f}".format(prop.B1p)
output_string += " {0:13.9f}".format(prop.B2p /
hybohr3togpa)
output_string += " {0:13.9f}".format(prop.B3p /
hybohr3togpa**2)
print(output_string)
pr = np.array(
[smin.Vmin, prop.E, prop.B, prop.B1p, prop.B2p, prop.B3p])
pmean = pmean + pr * pol[k1].w
pstd = pstd + (pr**2) * pol[k1].w
pstd = np.lib.scimath.sqrt(pstd - pmean**2)
if (LOG > 0):
print('\nAverage properties (weighted polynomials):')
print(
'------ ---volume-- ---energy-- --B-(GPa)-- ----B1p---- B2p-(1/GPa) B3p--(1/GPa^2)'
)
print(
'-mean- {0:11.6f} {1:11.6f} {2:11.6f} {3:11.6f} {4:11.6f} {5:14.9f}'
.format(pmean[0], pmean[1], pmean[2] * hybohr3togpa,
pmean[3], pmean[4] / hybohr3togpa,
pmean[5] / hybohr3togpa**2))
print(
'stdvev {0:11.6f} {1:11.6f} {2:11.6f} {3:11.6f} {4:11.6f} {5:14.9f}\n'
.format(float(np.real(pstd[0])), float(np.real(pstd[1])),
float(np.real(pstd[2])) * hybohr3togpa,
float(np.real(pstd[3])),
float(np.real(pstd[4])) / hybohr3togpa,
float(np.real(pstd[5])) / hybohr3togpa**2))
else:
# Analyze and report the reference geometry:
if (LOG > 0):
print(
'\n\navgpolyfit: AVERAGE OF {0} STRAIN POLYNOMIALS'.format(
strain))
print('Volume reference (V0): {0:.6f}'.format(V0))
print('Range of degrees: 2--{0}'.format(MaxDegree))
print('Number of polynomials: {0}'.format(npol))
print('\nProperties at the reference volume: {0}'.format(V0))
print(
'--i-- npol data order --SSerr-- ---w---- -----Eref--- ----Bref--- ---B1ref--- ---B2ref--- ---B3ref---'
)
isrt = np.argsort(ww)[::-1]
srt = ww[isrt]
pmean = np.zeros(6)
pstd = np.zeros(6)
for k in range(npol):
k1 = isrt[k]
prop = straineval(pol[k1].c, V0, V0, strain)
if (LOG > 0 and k <= 25):
# In the conversion of the bulk modulus and derivatives we
# assume the units: volume (bohr^3), energy (Hy).
output_string = "{0:4} {1:4} {2:4} {3:4}".format(
k, k1 + 1, pol[k1].data, pol[k1].order)
output_string += " {0:9.2e} {1:8.6f}".format(
pol[k1].SSerr, pol[k1].w)
output_string += " {0:13.6f}".format(prop.E)
output_string += " {0:10.6f}".format(prop.B *
hybohr3togpa)
output_string += " {0:9.6f}".format(prop.B1p)
output_string += " {0:13.9f}".format(prop.B2p /
hybohr3togpa)
output_string += " {0:13.9f}".format(prop.B3p /
hybohr3togpa**2)
print(output_string)
pr = np.array(
[V0, prop.E, prop.B, prop.B1p, prop.B2p, prop.B3p])
pmean = pmean + pr * pol[k1].w
pstd = pstd + (pr**2) * pol[k1].w
pstd = np.lib.scimath.sqrt(pstd - pmean**2)
if (LOG > 0):
print(
'\nAverage properties at the ref. volume: {0}'.format(V0))
print(
'------ ---energy-- --B-(GPa)-- ----B1p---- B2p-(1/GPa) B3p--(1/GPa^2)'
)
print(
'-mean- {0:11.6f} {1:11.6f} {2:11.6f} {3:11.6f} {4:14.9f}'.
format(pmean[1], pmean[2] * hybohr3togpa, pmean[3],
pmean[4] / hybohr3togpa,
pmean[5] / hybohr3togpa**2))
print(
'stdvev {0:11.6f} {1:11.6f} {2:11.6f} {3:11.6f} {4:14.9f}\n'
.format(float(np.real(pstd[1])),
float(np.real(pstd[2])) * hybohr3togpa,
float(np.real(pstd[3])),
float(np.real(pstd[4])) / hybohr3togpa,
float(np.real(pstd[5])) / hybohr3togpa**2))
Efit = strainevalE(cavg, V0, V, strain)
SSerr = np.sum((E - Efit)**2)
SStot = np.sum((E - np.mean(E))**2)
R2 = 1. - SSerr / SStot
if (nargout > 1):
savg = savg_class()
savg.eqmean = pmean
savg.eqstd = pstd
savg.R2 = R2
savg.Efit = Efit
#"""
if (nargout > 1):
# Putting error bars to the pressure, bulk modulus, etc at all volumes
savg.Emean, savg.Estd = np.zeros(len(V)), np.zeros(len(V))
savg.pmean, savg.pstd = np.zeros(len(V)), np.zeros(len(V))
savg.Bmean, savg.Bstd = np.zeros(len(V)), np.zeros(len(V))
savg.B1pmean, savg.B1pstd = np.zeros(len(V)), np.zeros(len(V))
savg.B2pmean, savg.B2pstd = np.zeros(len(V)), np.zeros(len(V))
savg.B3pmean, savg.B3pstd = np.zeros(len(V)), np.zeros(len(V))
for k in range(npol):
k1 = isrt[k]
prop = straineval(pol[k1].c, V0, V, strain)
savg.Emean = savg.Emean + prop.E * pol[k1].w
savg.Estd = savg.Estd + (prop.E**2) * pol[k1].w
savg.pmean = savg.pmean + prop.p * pol[k1].w
savg.pstd = savg.pstd + (prop.p**2) * pol[k1].w
savg.Bmean = savg.Bmean + prop.B * pol[k1].w
savg.Bstd = savg.Bstd + (prop.B**2) * pol[k1].w
savg.B1pmean = savg.B1pmean + prop.B1p * pol[k1].w
savg.B1pstd = savg.B1pstd + (prop.B1p**2) * pol[k1].w
savg.B2pmean = savg.B2pmean + prop.B2p * pol[k1].w
savg.B2pstd = savg.B2pstd + (prop.B2p**2) * pol[k1].w
savg.B3pmean = savg.B3pmean + prop.B3p * pol[k1].w
savg.B3pstd = savg.B3pstd + (prop.B3p**2) * pol[k1].w
savg.Estd = np.sqrt(savg.Estd - savg.Emean**2)
savg.pstd = np.sqrt(savg.pstd - savg.pmean**2)
savg.Bstd = np.sqrt(savg.Bstd - savg.Bmean**2)
savg.B1pstd = np.sqrt(savg.B1pstd - savg.B1pmean**2)
savg.B2pstd = np.sqrt(savg.B2pstd - savg.B2pmean**2)
savg.B3pstd = np.sqrt(savg.B3pstd - savg.B3pmean**2)
if nargout == 1:
return cavg
elif nargout == 2:
return cavg, savg
def point_cloud_to_panorama(points,
v_res=0.42,
h_res=0.35,
v_fov=(-24.9, 2.0),
d_range=(0, 100),
y_fudge=3):
""" Takes point cloud data as input and creates a 360 degree panoramic
image, returned as a numpy array.
Args:
points: (np array)
The numpy array containing the point cloud. .
The shape should be at least Nx3 (allowing for more columns)
- Where N is the number of points, and
- each point is specified by at least 3 values (x, y, z)
v_res: (float)
vertical angular resolution in degrees. This will influence the
height of the output image.
h_res: (float)
horizontal angular resolution in degrees. This will influence
the width of the output image.
v_fov: (tuple of two floats)
Field of view in degrees (-min_negative_angle, max_positive_angle)
d_range: (tuple of two floats) (default = (0,100))
Used for clipping distance values to be within a min and max range.
y_fudge: (float)
A hacky fudge factor to use if the theoretical calculations of
vertical image height do not match the actual data.
Returns:
A numpy array representing a 360 degree panoramic image of the point
cloud.
"""
# Projecting to 2D
x_points = points[:, 0]
y_points = points[:, 1]
z_points = points[:, 2]
r_points = points[:, 3]
d_points = np.sqrt(x_points**2 +
y_points**2) # map distance relative to origin
#d_points = np.sqrt(x_points**2 + y_points**2 + z_points**2) # abs distance
# We use map distance, because otherwise it would not project onto a cylinder,
# instead, it would map onto a segment of slice of a sphere.
# RESOLUTION AND FIELD OF VIEW SETTINGS
v_fov_total = -v_fov[0] + v_fov[1]
# CONVERT TO RADIANS
v_res_rad = v_res * (np.pi / 180)
h_res_rad = h_res * (np.pi / 180)
# MAPPING TO CYLINDER
x_img = np.arctan2(y_points, x_points) / h_res_rad
y_img = -(np.arctan2(z_points, d_points) / v_res_rad)
# THEORETICAL MAX HEIGHT FOR IMAGE
d_plane = (v_fov_total / v_res) / (v_fov_total * (np.pi / 180))
h_below = d_plane * np.tan(-v_fov[0] * (np.pi / 180))
h_above = d_plane * np.tan(v_fov[1] * (np.pi / 180))
y_max = int(np.ceil(h_below + h_above + y_fudge))
# SHIFT COORDINATES TO MAKE 0,0 THE MINIMUM
x_min = -360.0 / h_res / 2
x_img = np.trunc(-x_img - x_min).astype(np.int32)
x_max = int(np.ceil(360.0 / h_res))
y_min = -((v_fov[1] / v_res) + y_fudge)
y_img = np.trunc(y_img - y_min).astype(np.int32)
# CLIP DISTANCES
d_points = np.clip(d_points, a_min=d_range[0], a_max=d_range[1])
# CONVERT TO IMAGE ARRAY
img = np.zeros([y_max + 1, x_max + 1], dtype=np.uint8)
img[y_img, x_img] = scale_to_255(d_points, min=d_range[0], max=d_range[1])
return img
def trunc(a):
return np.trunc(a)
def direction(self):
"""
Perform and plot the two main directions of the peaks, considering their previously
calculated scale ,by calculating the Hessian at different sizes as the combination of
gaussians and their first and second derivatives
"""
import pylab
j = 0
vals = []
vects = []
kpx = self.keypoints.x
kpy = self.keypoints.y
sigma = self.keypoints.sigma
img = self.raw
pylab.figure()
pylab.imshow(img, interpolation='nearest')
for y, x, s in zip(kpy, kpx, sigma):
s_patch = numpy.trunc(s * 2)
if s_patch % 2 == 0:
s_patch += 1
if s_patch < 3: s_patch = 3
if (x > s_patch / 2 and x < img.shape[1] - s_patch / 2 - 1
and y > s_patch / 2 and y < img.shape[0] - s_patch / 2):
patch = img[y - (s_patch - 1) / 2:y + (s_patch - 1) / 2 + 1,
x - (s_patch - 1) / 2:x + (s_patch - 1) / 2 + 1]
x_patch = numpy.arange(s_patch)
Gx = numpy.exp(-4 * numpy.log(2) *
(x_patch - numpy.median(x_patch))**2 / s)
Gy = Gx[:, numpy.newaxis]
dGx = -Gx * 4 * numpy.log(2) / s * 2 * (x_patch -
numpy.median(x_patch))
dGy = dGx[:, numpy.newaxis]
d2Gx = -8 * numpy.log(2) / s * (
(x_patch - numpy.median(x_patch)) * dGx + Gx)
d2Gy = d2Gx[:, numpy.newaxis]
Hxx = d2Gx * Gy
Hyy = d2Gy * Gx
Hxy = dGx * dGy
d2x = (Hxx.ravel() * patch.ravel()).sum()
d2y = (Hyy.ravel() * patch.ravel()).sum()
dxy = (Hxy.ravel() * patch.ravel()).sum()
H = numpy.array([[d2y, dxy], [dxy, d2x]])
val, vect = numpy.linalg.eig(H)
# print 'new point'
# print x, y
# print val
# print vect
# print numpy.dot(vect[0],vect[1])
e = numpy.abs(val[0] - val[1]) / numpy.abs(val[0] + val[1])
j += 1
# print j
# print e
if numpy.abs(val[1]) < numpy.abs(
val[0]
): # reorganisation des valeurs propres et vecteurs propres
val[0], val[1] = val[1], val[0]
vect = vect[-1::-1, :]
pylab.annotate(
"",
xy=(x + vect[0][0] * val[0], y + vect[0][1] * val[0]),
xytext=(x, y),
arrowprops=dict(facecolor='red', shrink=0.05),
)
pylab.annotate(
"",
xy=(x + vect[1][0] * val[1], y + vect[1][1] * val[1]),
xytext=(x, y),
arrowprops=dict(facecolor='red', shrink=0.05),
)
pylab.plot(x, y, 'og')
vals.append(val)
vects.append(vect)
return vals, vects
def check_ltc_residuals(dataframe,
ccd,
rawy_range=(1, 200),
binned=None,
filter_select=None,
plot_it=True,
png_file=None,
title=''):
"""Validation for the LTC correction by means of checking the residuals
Parameters
----------
dataframe : dataframe, mandatory
the pandas dataframe with the Cu Kalpha fit results (from Michael Smith monitoring run). Produced by `ff_monitoring_work2.ipynb`
ccd : int, mandatory
the EPIC-pn CCD number (from 1 to 12)
rawy_range : list
the RAWY range to select, can be (1,200) and then can be (x,x+19) for x in range(1,201,20)
binned: float, optional
bin the data grouping with binned years, if None, then no binning
filter_select : str
if not None, then a selection on filter wheel is requested, can be one of 'CalClosed', 'CalMedium', 'CalThick', 'CalThin1', 'Closed',
'Medium', 'Thick', 'Thin1', 'Thin2'. If None, then all are selected.
plot_it : bool
if set, will plot the results.
png_file : str
is set, then the plotting results will be saved to this file.
title : str
Text to append to the end of the plot title, e.g. the version or other comment to apper on the plot title
Output
------
output: dict
{'ccd': []}
Method
------
Modification history
--------------------
Created 17 Mar 2021, Ivan Valchanov, XMM SOC
"""
#
ntot = np.count_nonzero((dataframe.ccd == ccd)
& (dataframe.rawy0 == rawy_range[0])
& (dataframe.rawy1 == rawy_range[1]))
#
xtab = select_data(dataframe,
ccd,
rawy_range=rawy_range,
filter_select=filter_select)
ntab = len(xtab)
xmode = xtab.xmode
if (filter_select is not None):
print(
f"CCD {ccd}, {xmode} mode, filter {filter_select}: filtered {ntab} results out of {ntot}"
)
else:
print(
f"CCD {ccd}, {xmode} mode: filtered {ntab} results out of {ntot}")
#
#line = dataframe.line
#
xin = xtab.delta_time
residual = (xtab.energy - line0) # in eV
residual_err = (xtab.energy_err1 + xtab.energy_err2) / 2.0 # in eV
qmean = np.mean(residual)
qstd = np.std(residual)
xstat = stats.sigma_clipped_stats(residual, sigma=3, maxiters=3)
#
if (binned is not None):
# add those as columns in the dataframe
qt = QTable.from_pandas(xtab)
qt['residual'] = residual
qt['residual_err'] = residual_err
#
year_bin = np.trunc(qt['delta_time'] / binned)
year_run = np.unique(year_bin)
#year_bin = np.trunc(qt['delta_time']/binned)
dat_grouped = qt.group_by(year_bin)
#
dat_binned = dat_grouped.groups.aggregate(np.median)
dat_binned_std = dat_grouped.groups.aggregate(mad)
xin_bin = dat_binned['delta_time']
yin_bin = dat_binned['residual']
yin_bin_err = dat_binned_std['residual']
#
# prepare the output
#
output = [
ccd, xmode, rawy_range[0], rawy_range[1], xstat[0], xstat[2], ntab,
year_run.data, xin_bin.data, yin_bin.data, yin_bin_err.data
]
#
# plotting
#
if (plot_it):
fig, ax = plt.subplots(figsize=(10, 6))
ax.errorbar(xin,
residual,
yerr=residual_err,
fmt='o',
label=f'CCDNR {ccd}',
zorder=0)
if (binned is not None):
ax.step(year_run,
yin_bin,
where='pre',
zorder=2,
color='cyan',
label='Per bin median')
#ax.step(xin_bin,yin_bin,where='mid',zorder=2,color='cyan',label='Per bin median')
ax.errorbar(xin_bin,
yin_bin,
yerr=yin_bin_err,
fmt='o',
color='cyan',
zorder=1)
ax.axhline(0.0, linestyle='dashed', linewidth=3, color='red', zorder=1)
ax.axhline(20.0,
linestyle='dotted',
linewidth=2,
color='red',
zorder=1)
ax.axhline(-20.0,
linestyle='dotted',
linewidth=2,
color='red',
zorder=1)
ax.text(0.1,
0.9,
fr'mean={qmean:.1f} eV, st.dev.={qstd:.1f} eV',
fontsize=14,
transform=ax.transAxes)
ax.text(
0.1,
0.8,
fr'mean={xstat[0]:.1f} eV, st.dev.={xstat[2]:.1f} eV (3-$\sigma$ clipped)',
fontsize=14,
transform=ax.transAxes)
ax.set_xlim((0.0, 22.0))
ax.set_ylim((-100.0, 100.0))
ax.grid(True)
ax.legend(loc=3)
ax.set_title(
f"Cu-Ka data for EPIC-PN CCD{ccd:02}, mode={xmode}, RAWY in [{rawy_range[0]},{rawy_range[1]}], {title}"
)
ax.set_ylabel(r"E$_{corr}$ - E$_{lab}$ (eV)")
ax.set_xlabel("Time since 2000-01-01 (years)")
if (png_file is not None):
plt.savefig(png_file, dpi=100)
plt.show()
plt.close()
return output
def set_fold(self):
fc_mins = np.trunc(np.min(self.log2fc, axis=0))
fc_maxs = np.trunc(np.max(self.log2fc, axis=0))
fold = np.stack([fc_mins, fc_maxs])
return fold
lg = np.zeros((rN, rN, rN, 3))
local_mass = 0.0e0
local_g = 0.0e0
local_v = 0.0e0
local_sigma0k = np.zeros((rN, rN, rN, 3, 3))
local_sigmav = np.zeros((rN, rN, rN, 3, 3))
for t in range(fhstep / atmstep):
print('t;', t)
t_ = fhstep / atmstep * it + t
pos = infh['coordinates'][t_]
vel = infh['velocities'][t_]
box = infh['cell_lengths']
L = box[0, 0]
rN = int(L / cutoff)
# unset PBC
pos -= np.trunc(pos / L) * L
pos += np.round((0.50e0 * L - pos) / L) * L
pos = cp.asarray(pos, dtype=np.float32)
# grid parameters
r_min = 0.0e0
r_max = L
dr = (r_max - r_min) / float(rN)
dV = dr**3
if t == 0:
r_ = np.array([r_min + dr * ir for ir in range(rN)])
rax = np.array([r_, r_, r_])
# calculate rho, g, v, sigma_ab, tau_ab, S_ab, eta as local variables
m = cp.asarray(masses, dtype=np.float32)
lm = cp.zeros((rN, rN, rN), dtype=cp.float32)
r0 = cp.asarray(pos[:, 0] / dr, dtype=cp.int32)
import numpy as np
import cv2
import pandas as pd
import dataframe
#下载数据
data_all=np.loadtxt("Image_all.txt")
data_i=np.unique(data_all[:,0])
print(data_i)
k = 0
data_i = np.array(data_i)
data_i = np.trunc(data_i)
data_i = data_i.astype('int')
print(len(data_i))
for i in range(len(data_i)):
data=[]
for i in range(len(data_all)):
if(data_i[k]==data_all[i,0]):
data.append(data_all[i])
else:
continue
k=k+1
data=np.array(data)
data=np.trunc(data)
data=data.astype('int')
data3=data.copy()
#找多个元素是否同行的循环次数
# s=0
# 按第3列X0进行排序，按高度排列
data = data[data[:, 2].argsort()]
# -*- coding: utf-8 -*-
from __future__ import unicode_literals
import numpy as np
a = np.array([20, 20, -20, -20])
b = np.array([3, -3, 6, -6])
# 真除
c = np.true_divide(a, b)
c = np.divide(a, b)
c = a / b
print('array:', c)
# 对ndarray做floor操作
d = np.floor(a / b)
print('floor_divide:', d)
# 对ndarray做ceil操作
e = np.ceil(a / b)
print('ceil ndarray:', e)
# 对ndarray做trunc操作
f = np.trunc(a / b)
print('trunc ndarray:', f)
# 对ndarray做around操作
g = np.around(a / b)
print('around ndarray:', g)
signif_clusters = [clusters[s] for s in signif]
signif_cluster_pvals = cluster_pvals[signif]
# plot stats
for clu, pv in zip(signif_clusters, signif_cluster_pvals):
'''
# this index tells direction of tval, hence could be used to
# decide which color to draw the significant cluster region
# based on which curve is higher:
idx = (np.sign(tvals[clu[0][0], 0]).astype(int) + 1) // 2
'''
clu = clu[0]
cluster_ymin = ylim[0] * np.ones_like(t[clu])
cluster_ymax = np.max(contr_mean[:, clu], axis=0)
pval_x = t[int(np.mean(clu[[0, -1]]))]
pval_y = -0.1 * ylim[1]
pval_ord = np.trunc(np.log10(pv)).astype(int)
_ = axs[ii].fill_between(t[clu], cluster_ymin, cluster_ymax,
alpha=1, facecolor='0.9', zorder=1,
edgecolor='none')
if show_pval:
pval_txt = '$p < 10^{{{}}}$'.format(pval_ord)
_ = axs[ii].text(pval_x, pval_y, pval_txt, ha='center',
va='baseline', fontdict=dict(size=10))
# set axis limits
_ = axs[ii].set_ylim(*ylim)
_ = axs[ii].set_xlim(*xlim)
# remove yaxis / ticks / ticklabels near bottom
ytck = [-0.1 * ymax, ymax]
ytl = axs[ii].yaxis.get_ticklocs()
_ = axs[ii].spines['left'].set_bounds(*ytck)
_ = axs[ii].yaxis.set_ticks(ytl[ytl > ytck[0]])
plt.ylabel('True label')
plt.xlabel('Predicted label')
classes = ['A-antigen negative', 'A-antigen positive']
plt.grid('off')
plt.colorbar()
tick_marks = np.arange(len(classes))
plt.xticks(tick_marks, classes, rotation=45)
plt.yticks(tick_marks, classes)
for i in range(cnf_matrix.shape[0]):
for j in range(cnf_matrix.shape[1]):
plt.text(j,
i,
cnf_matrix[i, j],
horizontalalignment="center",
color="orangered")
plt.gcf().subplots_adjust(left=0.25, bottom=0.35)
plt.savefig('Images/A_ConfusionWeighed.png', format='png', dpi=300)
coefPaths = justVarPathsNew[idxNZ[1]]
tile_path = np.trunc(coefPaths / (16**5))
tile_step = np.trunc((coefPaths - tile_path * 16**5) / 2)
tile_phase = np.trunc((coefPaths - tile_path * 16**5 - 2 * tile_step))
vtile_path = vhex(tile_path.astype('int'))
vitle_step = vhex(tile_step.astype('int'))
def adaPoint(box, pro):
box_pro = box
if pro != 1.0:
box_pro = box / pro
box_pro = np.trunc(box_pro)
return box_pro
'''
Created on 2019年8月18日
@author: lingyulong
'''
import numpy as np
#np.abs():每个元素取绝对值
data = np.array([1, 2, -3, 4, -5, 6])
print("原数组:", data)
print("数组进行abs操作:", np.abs(data))
data = np.array([[1.1, 1.5, 1.7], [-1.1, -1.5, -1.7], [2.1, 2.6, 3.9]])
print(np.rint(data)) #对每个元素进行四舍五入
print(np.trunc(data)) #对每个元素进行向0取整
print(np.isinf(data))
print(np.isfinite(data))
print(np.isnan(data))
#np.add():两个数组的对应元素相加
a = np.array([[1, 2, 3], [11, 22, 33]])
b = np.array([[1, -1, 2], [2, 3, 1]])
print("两个数组对应元素相加得到的新数组：")
c = np.add(a, b)
print(c)
#np.maximum():两个数组的对应元素相加
a = np.array([[1, 2, 3], [11, 22, 33]])
b = np.array([[1, 15, 9], [2, 3, 100]])
print("两个数组对应元素的最大值/最小值：")
print(np.maximum(a, b))
def plot_gain_circles(self, two_port, gains_db=None, plane='source', \
gain_cursor=True, surface=False):
self.two_port = two_port
self.gain_cursor = gain_cursor
self.current_plane = plane
max_gain = self.two_port.g_max()
if (self.two_port.kt() < 1).any():
max_gain = self.two_port.max_double_sided_mismatched_gain()
if gains_db == None:
gains_db = np.trunc(
10 * db(max_gain)[0] * linspace(0.5, 1, 4)) / 10
if surface:
filled = False
circle_alpha = 0.7
linestyle = 'dashed'
else:
filled = True
circle_alpha = 0.2
linestyle = 'solid'
for g in gains_db:
if plane == 'source':
center, radius = self.two_port.available_gain_circle(un_db(g))
self.ax.set_title('Source plane')
elif plane == 'load':
center, radius = self.two_port.operating_gain_circle(un_db(g))
self.ax.set_title('Load plane')
text = str(g) + ' dB'
self.plot_circle((center[0].real, center[0].imag), radius[0], text=text, filled=filled,\
text_color='black', circle_color='orange', circle_alpha=circle_alpha,\
linestyle=linestyle)
if not surface:
self.save_background()
return
num_segments = 1000
i, r = meshgrid(linspace(-1.1, 1.1, num_segments),
linspace(-1.1, 1.1, num_segments))
gamma = i + 1j * r
term = OnePort(s=gamma.reshape(-1))
if plane == 'source':
g = self.two_port.g_a(term)
elif plane == 'load':
g = self.two_port.g_p(term)
g = db(g).reshape(num_segments, num_segments)
g[abs(gamma) > 1] = nan
im = self.ax.imshow(g,
origin='lower',
extent=(-1.1, 1.1, -1.1, 1.1),
interpolation='bicubic',
alpha=0.9)
im.set_clim(gains_db[0], db(max_gain))
self.save_background()
def init(self):
global b1
#devel06
b1_fname='/home/ian/Data/dipy/brain2_scan1_fiber_track_mni.trk'
#devel07
#b1_fname='/home/eg01/Data_Backup/Data/PBC/pbc2009icdm/brain2/brain2_scan1_fiber_track_mni.trk'
b1=tracks.Tracks(b1_fname,subset=[0,200])
b1.angular_speed = 0.
b1.picking_example = True
b1.min_length = 20.
b1.opacity = 0.8
b1.manycolors = False
b1.brain_color = [0, 0, 0]
b1.init()
global texim
#devel06
fname = '/home/ian/Data/dipy/Streaks4.bmp'
#devel07
#fname = '/home/eg01/Devel/Fos/fos/core/tests/data/Streaks4.bmp'
texim = texture.Texture_Demo(fname,red=False,green=False, blue=True)
#texim.orbit = b1.data[4246]
#print 'len(b1.data)', len(b1.data)
random_inx=np.trunc(len(b1.data)*np.random.rand(200)).astype(np.int)
#print random_inx
#texim.orbits = [b1.data[4246],b1.data[3000],b1.data[2000],b1.data[1000]]
texim.orbits =[]
for i in random_inx:
#print i
if tm.length(b1.data[i]) > 20.:
#print i
texim.orbits.append(b1.data[i])
texim.orbits_index = np.zeros((len(texim.orbits),),np.int)
texim.init()
self.slots={0:{'actor':texim,'slot':( 0, 800*MS )}}#,
def trunc(x, ndecimals=0):
"""Truncate a floating point number to a given number of decimals."""
decade = 10**ndecimals
return np.trunc(x * decade) / decade
def tax_of_payment_confirmation_money(self, engineer_history):
if engineer_history:
return numpy.trunc((self.payment_confirmation_money or 0) * engineer_history.payment_tax.rate)
else:
return 0
