def sub_retro(rx, ry, sx, sy, times):
"""Creates a 2 by 3 subplot of the system at 6 chosen times
Parameters
----------
rx: array, x postitions of all stars
ry: array, y postitions of all stars
sx: array, x postitions of S
sy: array, y postitions of S
times: float, time at which you are plotting
Returns
-------
2 by 3 subplot of the system at 6 chosen times
"""
fig, ax = plt.subplots(2,3, figsize=(12,8))
c=1
for i in range(2):
for j in range(3):
plt.sca(ax[i,j])
t=times[i,j]
plt.scatter(0,0,color= 'c', label= 'M', s=40)
plt.scatter(sx[t], sy[t], color = 'b', label = 'S', s=40)
plt.scatter(rx[t], ry[t], color = 'g', label = 'stars', s=7)
plt.xlabel(c+2, fontsize=15)
plt.xlim(-45,45)
plt.ylim(-45,45)
plt.box(False)
plt.tick_params(axis = 'x', top = 'off', bottom = "off", labelbottom= 'off')
plt.tick_params(axis = 'y', right = 'off', left= "off", labelleft= 'off')
c+=1
plt.tight_layout()
def init_plot(self):
self.get_subject_data()
self.compute_bounds()
self.compute_confidence_mapping()
self.compute_predition()
plt.sca(self.rt_ax)
plt.step(self.rt_centers,self.subject_rt[0],'b',label='Subject hit')
plt.step(self.rt_centers,-self.subject_rt[1],'r',label='Subject miss')
self.hit_rt, = plt.plot(self.model_t,self.model_rt[0],'b',label='Model hit',linewidth=2)
self.miss_rt, = plt.plot(self.model_t,-self.model_rt[1],'r',label='Model hit',linewidth=2)
self.rt_ax.set_xlim([0,self._max_RT])
plt.xlabel('RT')
plt.ylabel('Prob density')
plt.legend(loc='best', fancybox=True, framealpha=0.5)
plt.sca(self.conf_ax)
plt.step(self.c_centers,self.subject_conf[0],'b',label='Subject hit')
plt.step(self.c_centers,-self.subject_conf[1],'r',label='Subject miss')
self.hit_conf, = plt.plot(np.linspace(0,1,self.model_conf.shape[1]),self.model_conf[0],'b',label='Model hit',linewidth=2)
self.miss_conf, = plt.plot(np.linspace(0,1,self.model_conf.shape[1]),-self.model_conf[1],'r',label='Model hit',linewidth=2)
plt.xlabel('RT')
plt.ylabel('Prob density')
plt.legend(loc='best', fancybox=True, framealpha=0.5)
self.title = plt.suptitle('full_nLL = {full_nLL}\nconfidence_only_nLL = {confidence_only_nLL}\nfull_confidence_nLL = {full_confidence_nLL}\n'.format(
full_nLL=self.full_nLL,confidence_only_nLL=self.confidence_only_nLL,full_confidence_nLL=self.full_confidence_nLL))
def stick_bgplvm(model=None, optimize=True, verbose=True, plot=True):
"""Interactive visualisation of the Stick Man data from Ohio State University with the Bayesian GPLVM."""
from GPy.models import BayesianGPLVM
from matplotlib import pyplot as plt
import numpy as np
import GPy
import pods
data = pods.datasets.osu_run1()
Q = 6
kernel = GPy.kern.RBF(Q, lengthscale=np.repeat(.5, Q), ARD=True)
m = BayesianGPLVM(data['Y'], Q, init="PCA", num_inducing=20, kernel=kernel)
m.data = data
m.likelihood.variance = 0.001
# optimize
try:
if optimize: m.optimize('bfgs', messages=verbose, max_iters=5e3, bfgs_factor=10)
except KeyboardInterrupt:
print("Keyboard interrupt, continuing to plot and return")
if plot:
fig, (latent_axes, sense_axes) = plt.subplots(1, 2)
plt.sca(latent_axes)
m.plot_latent(ax=latent_axes)
y = m.Y[:1, :].copy()
data_show = GPy.plotting.matplot_dep.visualize.stick_show(y, connect=data['connect'])
dim_select = GPy.plotting.matplot_dep.visualize.lvm_dimselect(m.X.mean[:1, :].copy(), m, data_show, latent_axes=latent_axes, sense_axes=sense_axes)
fig.canvas.draw()
# Canvas.show doesn't work on OSX.
#fig.canvas.show()
raw_input('Press enter to finish')
return m
def plot_some_shots(run, ax=None):
file_name = h5_file_name_funk(run)
h5 = load_file(file_name)
spec_group = h5['spectral_properties']
#####
# Select region in the tree to get traces
# centers = (spec_group['center_eV'].value -
# h5['photoelectron_energy_prediction_eV'].value)
widths = spec_group['width_eV'].value
#I = np.abs(centers) < (center_max - center_min) / 10
# I = ((np.nanmax(widths) - (np.nanmax(widths) - np.nanmin(widths)) / 5) <
# widths)
I = np.isfinite(widths)
selected = np.random.choice(np.where(I)[0], 5)
selected.sort()
energy_scale_eV = h5['energy_scale_eV'].value
traces = h5['energy_signal'].value
if ax is None:
plt.figure('energy traces ' + str(run))
plt.clf()
else:
plt.sca(ax)
for shot in selected:
plt.plot(energy_scale_eV, traces[shot, :] * 1e3)
plt.xlim(70, 130)
def plotBestFit(weights):
m = shape(dataMat)[0]
xcord1 = []
ycord1 = []
xcord2 = []
ycord2 = []
for i in range(m):
if labelMat[i] == 1:
xcord1.append(dataMat[i, 1])
ycord1.append(dataMat[i, 2])
else:
xcord2.append(dataMat[i, 1])
ycord2.append(dataMat[i, 2])
plt.figure(1)
ax = plt.subplot(111)
ax.scatter(xcord1, ycord1, s=30, c='red', marker='s')
ax.scatter(xcord2, ycord2, s=30, c='green')
x = arange(0.2, 0.8, 0.1)
y = array((-weights[0] - weights[1] * x) / weights[2])
print shape(x)
print shape(y)
plt.sca(ax)
plt.plot(x, y) # ramdomgradAscent
# plt.plot(x,y[0]) #gradAscent
plt.xlabel('density')
plt.ylabel('ratio_sugar')
# plt.title('gradAscent logistic regression')
plt.title('ramdom gradAscent logistic regression')
plt.show()
def _plot_cc_mat(pdata, ax):
plt.sca(ax)
plt.imshow(pdata['CC'], interpolation='nearest', aspect='auto', vmin=-0.5, vmax=0.5, cmap=plt.cm.seismic)
plt.yticks(range(nelectrodes), ['%d' % _e for _e in electrode_order])
plt.xticks(range(nfreqs), ['%d' % _f for _f in freqs])
plt.title(pdata['aprop'])
plt.colorbar()
def demo_alternatinggraphcut():
t = 12
## Simulate small image
im, G = simulation.simulate_image(11,13,t,theta=0.4*np.pi )
# Show grid overlaid image
fig, ax = plt.subplots(1,2)
plt.sca(ax[0])
plt.imshow(im,cmap=plt.cm.gray)
plt.title('True grid')
plotgrid(G)
# Add random, disturbing, points
Nnew = 15
x_new = np.random.uniform(np.min(G.xy[:,0]),np.max(G.xy[:,0]), Nnew)
y_new = np.random.uniform(np.min(G.xy[:,1]),np.max(G.xy[:,1]), Nnew )
xy_new = np.vstack((x_new,y_new)).T
# Add to grid and make naive guess for edges
G.xy = np.vstack((G.xy,xy_new))
G.resolve_edges(remove_long=False)
# Show this modified grid
plt.sca(ax[1])
plotgrid(G)
plt.title('Modified/disturbed grid')
plt.show()
# Clean up using alternating graph cut
xy_hat, simplices_hat = alternating_graphcut.cleanup(G.xy,im,alpha=3,beta=2)
G_hat = grid.TriangularGrid.from_simplices(xy_hat,simplices_hat)
plt.figure()
plotgrid(G_hat)
plt.show(block=True)
def map(self, func, **kwargs):
"""Plot with the same function in every subplot.
Parameters
----------
func : callable plotting function
Must take x, y arrays as positional arguments and draw onto the
"currently active" matplotlib Axes.
"""
kw_color = kwargs.pop("color", None)
for i, y_var in enumerate(self.y_vars):
for j, x_var in enumerate(self.x_vars):
hue_grouped = self.data.groupby(self.hue_vals)
for k, (label_k, data_k) in enumerate(hue_grouped):
ax = self.axes[i, j]
plt.sca(ax)
color = self.palette[k] if kw_color is None else kw_color
func(data_k[x_var], data_k[y_var],
label=label_k, color=color, **kwargs)
self._clean_axis(ax)
self._update_legend_data(ax)
if kw_color is not None:
kwargs["color"] = kw_color
self._add_axis_labels()
def map_upper(self, func, **kwargs):
"""Plot with a bivariate function on the upper diagonal subplots.
Parameters
----------
func : callable plotting function
Must take x, y arrays as positional arguments and draw onto the
"currently active" matplotlib Axes.
"""
kw_color = kwargs.pop("color", None)
for i, j in zip(*np.triu_indices_from(self.axes, 1)):
hue_grouped = self.data.groupby(self.hue_vals)
for k, (label_k, data_k) in enumerate(hue_grouped):
ax = self.axes[i, j]
plt.sca(ax)
x_var = self.x_vars[j]
y_var = self.y_vars[i]
color = self.palette[k] if kw_color is None else kw_color
func(data_k[x_var], data_k[y_var], label=label_k,
color=color, **kwargs)
self._clean_axis(ax)
self._update_legend_data(ax)
if kw_color is not None:
kwargs["color"] = kw_color
def colorline(x, y, z=None, cmap=plt.get_cmap('copper'), ax=None, norm=plt.Normalize(0.0, 1.0),
label='', linewidth=3, alpha=1.0):
'''
Plot a colored line with coordinates x and y
Optionally specify colors in the array z
Optionally specify a colormap, a norm function and a line width
'''
from matplotlib.collections import LineCollection
# Default colors equally spaced on [0,1]:
if z is None:
z = np.linspace(0.0, 1.0, len(x))
# Special case if a single number:
if not hasattr(z, "__iter__"): # to check for numerical input -- this is a hack
z = np.array([z])
z = np.asarray(z)
if ax is None:
fig, ax = plt.subplots(nrows=1, ncols=1)
fig.patch.set_facecolor('white')
else:
plt.sca(ax)
fig = plt.gcf()
segments = make_segments(x, y)
lc = LineCollection(segments, array=z, cmap=cmap, norm=norm, linewidth=linewidth, alpha=alpha)
ax.add_collection(lc)
return lc
def plot_marginals(self, func, **kwargs):
"""Draw univariate plots for `x` and `y` separately.
Parameters
----------
func : plotting callable
This must take a 1d array of data as the first positional
argument, it must plot on the "current" axes, and it must
accept a "vertical" keyword argument to orient the measure
dimension of the plot vertically.
kwargs : key, value mappings
Keyword argument are passed to the plotting function.
Returns
-------
self : JointGrid instance
Returns `self`.
"""
plt.sca(self.ax_marg_x)
func(self.x, **kwargs)
kwargs["vertical"] = True
plt.sca(self.ax_marg_y)
func(self.y, **kwargs)
return self
def add_bovy_rix(axScatter,Rmin,Rmax,zmin,zmax):
#Now calculate and plot our model
fehs= numpy.linspace(-1.6,0.5,26)
afes= numpy.linspace(-0.15,0.55,25)
ourDist= numpy.zeros((len(fehs),len(afes)))
for ii in range(len(fehs)):
for jj in range(len(afes)):
ourDist[ii,jj]= maps.abundanceDist(fehs[ii],afes[jj],
z=[zmin*1000.,zmax*1000.],
r=[Rmin,Rmax])
_ADJUSTABUNDANCES= True
_ADJUSTFEH= True
if _ADJUSTABUNDANCES:
if _ADJUSTFEH:
fehs+= 0.15
afes/= 1.5
afes-= 0.025
#Contour this in axScatter
ourDist[numpy.isnan(ourDist)]= 0.
sortindx= numpy.argsort(ourDist.flatten())[::-1]
cumul= numpy.cumsum(numpy.sort(ourDist.flatten())[::-1])/numpy.sum(ourDist.flatten())
cntrThis= numpy.zeros(numpy.prod(ourDist.shape))
cntrThis[sortindx]= cumul
cntrThis= numpy.reshape(cntrThis,ourDist.shape)
pyplot.sca(axScatter)
CS= pyplot.contour(fehs,afes,
cntrThis.T,levels=[0.68,0.95],
linewidths=3.,linestyles='dashed',
colors='r',zorder=10)
def plot_locations(self, ax=None, lu=None):
if self.annotations_loaded == False:
return
if ax is None:
fig, ax = pl.subplots(1, 1, sharex=True, sharey=False, figsize=(20, 5))
else:
pl.sca(ax)
if lu is None:
lu = (self.meta['start'], self.meta['end'])
palette = it.cycle(sns.husl_palette())
offsets = self.get_offsets()
for ai in xrange(self.num_annotators):
col = next(palette)
offset = offsets[ai]
for index, rr in slice_df_start_stop(self.locations[ai], lu).iterrows():
pl.plot([rr['start'], rr['end']], [self.location_targets.index(rr['name']) + offset * 2] * 2, color=col,
linewidth=5, alpha=0.5)
pl.yticks(np.arange(len(self.location_targets)), self.location_targets)
pl.ylim((-1, len(self.location_targets)))
pl.xlim(lu)
def plot_session_PSTH(self, session, tetrode, experiment=-1, site=-1, cluster = None, sortArray='currentFreq', timeRange = [-0.5, 1], replace=0, lw=3, colorEachCond=None):
sessionObj = self.get_session_obj(session, experiment, site)
sessionDir = sessionObj.ephys_dir()
ephysData, bdata, info = self.load_session_data(session, experiment, site, tetrode, cluster)
eventOnsetTimes = ephysData['events']['stimOn']
spikeTimestamps = ephysData['spikeTimes']
if bdata is not None:
sortArray = bdata[sortArray]
if colorEachCond is None:
colorEachCond = self.get_colours(len(np.unique(sortArray)))
else:
sortArray = []
plotTitle = info['sessionDir']
ephysData = ephyscore.load_ephys(sessionObj.subject, sessionObj.paradigm, sessionDir, tetrode, cluster)
eventOnsetTimes = ephysData['events']['stimOn']
spikeTimestamps = ephysData['spikeTimes']
if replace==1:
plt.cla()
elif replace==2:
plt.sca(ax)
else:
plt.figure()
plot_psth(spikeTimestamps, eventOnsetTimes, sortArray = sortArray, timeRange=timeRange, lw=lw, colorEachCond=colorEachCond, plotLegend=0)
def niceplot(ax=None, axfs='12', lfs='14', tightlayout=True,
mew=1.25, lw=2.0, ms=7.0, **kwargs):
"""Pretty up a plot for publication.
Parameters
----------
ax : matplotlib.axes.AxesSubplot, optional
An axis to niceify. Default is all axes in the current figure.
axfs : string, float, or int, optional
Axis tick label font size.
lfs : string, float, or int, optional
Axis label font size.
tightlayout : bool, optional
Run `plt.tight_layout`.
**kwargs
Any line or marker property keyword.
"""
import matplotlib.pyplot as plt
if ax is None:
for ax in plt.gcf().get_axes():
niceplot(ax, tightlayout=tightlayout, axfs=axfs, lfs=lfs, **kwargs)
# for the axes
plt.setp(ax.get_ymajorticklabels(), fontsize=axfs)
plt.setp(ax.get_xmajorticklabels(), fontsize=axfs)
# axis labes
labels = (ax.xaxis.get_label(), ax.yaxis.get_label())
plt.setp(labels, fontsize=lfs)
# for plot markers, ticks
lines = ax.get_lines()
mew = kwargs.pop('markeredgewidth', kwargs.pop('mew', None))
if mew is not None:
plt.setp(lines, mew=mew)
ms = kwargs.pop('markersize', kwargs.pop('ms', None))
if ms is not None:
plt.setp(lines, ms=ms)
lw = kwargs.pop('linewidth', kwargs.pop('lw', None))
if lw is not None:
plt.setp(lines, lw=lw)
if len(kwargs) > 0:
plt.setp(lines, **kwargs)
lines = ax.xaxis.get_minorticklines() + ax.xaxis.get_majorticklines() + \
ax.yaxis.get_minorticklines() + ax.yaxis.get_majorticklines()
plt.setp(lines, mew=1.25)
# the frame
plt.setp(ax.patch, lw=2.0)
if hasattr(plt, "tight_layout") and tightlayout:
plt.sca(ax)
plt.tight_layout()
def plot2d(signal, tmin=0.0, tmax=None, axes=None, **kwargs):
plot_type = kwargs.pop('type', 'contourf')
nlevels = int(kwargs.pop('nlevels', 100))
default_min = float(kwargs.pop('minrange', 0.))
default_max = float(kwargs.pop('maxrange', 1.))
plot_func = getattr(plt, plot_type)
xaxis = getattr(signal, signal.axes[1])
yaxis = getattr(signal, signal.axes[0])
xaxis[:]
yaxis[:]
if tmax is None:
tmax = yaxis[-1] # this should be yaxis.max() when fixed
plot_range = set_range(signal, default_min, default_max)
levels = np.linspace(plot_range[0], plot_range[1], nlevels)
plt.sca(axes)
artist = plot_func(np.array(xaxis), np.array(yaxis), np.array(signal),
levels=levels, **kwargs)
plt.ylabel('{} ({})'.format(yaxis._name, yaxis.units))
plt.xlabel('{} ({})'.format(xaxis._name, xaxis.units))
plt.title('{} -- {} -- {}'.format(signal._parent._name.upper(),
signal._name,
signal.shot))
plt.ylim(tmin, tmax)
if plot_type == 'contourf':
cbar = plt.colorbar(artist, format='%.1e')
cbar.set_label(signal.units, rotation=270)
def plot(w):
dataMat=np.array(df[['density','ratio_sugar']].values[:,:])
labelMat=np.mat(df['good'].values[:]).transpose()
m=np.shape(dataMat)[0]
xcord1=[]
ycord1=[]
xcord2=[]
ycord2=[]
for i in range(m):
if labelMat[i]==1:
xcord1.append(dataMat[i,0])
ycord1.append(dataMat[i,1])
else:
xcord2.append(dataMat[i,0])
ycord2.append(dataMat[i,1])
plt.figure(1)
ax=plt.subplot(111)
ax.scatter(xcord1,ycord1,s=30,c='red',marker='s')
ax.scatter(xcord2,ycord2,s=30,c='green')
x=np.arange(-0.2,0.8,0.1)
y=np.array((-w[0,0]*x)/w[0,1])
plt.sca(ax)
plt.plot(x,y) #gradAscent
plt.xlabel('density')
plt.ylabel('ratio_sugar')
plt.title('LDA')
plt.show()
def plot():
plt.ioff()
fig_mpl, ax = plt.subplots(1,figsize=(8,8), facecolor='white');
plt.xlim(-75,75);
plt.ylim(-75,75);
plt.sca(ax);
ax.set_xlabel("X Position (kpc)");
ax.set_ylabel("Y Position (kpc)");
plt.title("Relative Motion");
plt.legend()
scat = ax.scatter(X[0],Y[0],color='red',label='Disrupting Galaxy, Mass S',s=10);
scatt = ax.scatter(x,y,color='green',s=5);
scattt = ax.scatter(0,0,color='black',label='Central Galaxy, Mass M',s=10);
def make_gif(u):
newX,newY = X[u*40],Y[u*40]
scat.set_offsets(np.transpose(np.vstack([newX,newY])))
newx=[x[k][u*40] for k in range(120)]
newy=[y[k][u*40] for k in range(120)]
scatt.set_offsets(np.transpose(np.vstack([newx,newy])))
return mplfig_to_npimage(fig_mpl)
if parabolic == 1:
animation = mpy.VideoClip(make_gif,duration=tmax*25);
else:
animation = mpy.VideoClip(make_gif,duration=tmax*5);
#animation.ipython_display(fps=50)
animation.write_videofile("Retrograde Elliptic Passage.mp4", fps=50)
def create_pseudo_random_code(clen=10000,rseed=0,verbose=False):
"""
Create waveform files for hfradar
Juha Vierinen
"""
Npt = 200 # number of points to plot, just for plotting, arbitrary
"""
seed is a way of reproducing the random code without having to store all actual codes.
the seed can then act as a sort of station_id.
"""
seed(rseed)
"""
generate a uniform random phase modulated (complex) signal 'sig".
It's single precision floating point for SDR, since DAC is typically <= 16 bits!
"""
sig = np.exp(1j*2.0*np.pi*random(clen)).astype('complex64')
if stuffr is not None:
stuffr.plot_cts(sig[:Npt])
if verbose and hist is not None:
fg,ax = subplots(3,1)
sca(ax[0])
hist(sig.real)#,50)
sca(ax[1])
hist(sig.imag)
#hist(random(clen))
return sig
def plot_col_pos(cs_xzdf, cs_xydf, cs_xdf, colposperiod, colpostype):
dat=pd.date_range(start=cs_xzdf.index[0],end=cs_xzdf.index[-1], freq=colposperiod)
fig,ax=plt.subplots(nrows=1,ncols=2, sharex=True, sharey=True,figsize=fig_size)
plt.suptitle(colname+"\n"+colpostype+" col pos")
cm = plt.get_cmap('gist_rainbow')
ax[0].set_color_cycle([cm(1.*(len(dat)-i-1)/len(dat)) for i in range(len(dat))])
ax[1].set_color_cycle([cm(1.*(len(dat)-i-1)/len(dat)) for i in range(len(dat))])
for d in range(len(dat)):
curxz=cs_xzdf[(cs_xzdf.index==dat[d])]
curxy=cs_xydf[(cs_xydf.index==dat[d])]
curx=cs_xdf[(cs_xdf.index==dat[d])]
if colpostype=='rel':
curxz=curxz.sub(cs_xzdf.iloc[0,:],axis=1)
curxy=curxy.sub(cs_xydf.iloc[0,:],axis=1)
plt.sca(ax[0])
plt.axis('equal')
plt.plot([[0]]+curxz.values.T.tolist(),[[0.0]]+curx.values.T.tolist(), '.-')
plt.sca(ax[1])
plt.axis('equal')
plt.plot([[0.0]]+curxy.values.T.tolist(),[[0.0]]+curx.values.T.tolist(), '.-',label=datetime.strftime(dat[d],'%Y-%m-%d'))
ax[0].set_xlabel("XZ disp, m \n (+) downslope", fontsize='small',horizontalalignment='center')
ax[0].set_ylabel("X disp, m \n (+) towards surface", fontsize='small', rotation='vertical',horizontalalignment='center')
ax[1].set_xlabel("XY disp, m \n (+) to the right, facing downslope", fontsize='small', horizontalalignment='center')
plt.legend(loc='lower right', fontsize='small')
plt.tight_layout()
plt.subplots_adjust(left=None, bottom=None, right=None, top=0.9,wspace=None, hspace=None)
return fig,ax
def scale_axis_limits( self, k1, k2, k3, k4 ):
# k1, k2, k3, k4 are scaling parameters for
# xmin, xmax, ymin, ymax, respectively.
sca( self.ax )
[xmin, xmax, ymin, ymax] = axis()
self.ax.set_xlim( xmin * k1, xmax * k2 )
self.ax.set_ylim( ymin * k3, ymax * k4 )
def labeledDendrogram(dmat, labels, method='complete', cmap=None):
"""Perform hierarchical clustering on df columns and plot square heatmap of pairwise distances"""
"""TODO: add tick labels, with sparsity option"""
Z = sch.linkage(dmat, method=method)
den = sch.dendrogram(Z, color_threshold=np.inf, no_plot=True)
figh = plt.gcf()
figh.clf()
denAX = figh.add_axes([0.32, 0.05, 0.6, 0.9])
cbAX = figh.add_axes([0.25, 0.05, 0.05, 0.9])
plt.sca(denAX)
denD = sch.dendrogram(Z, color_threshold=np.inf, orientation='left')
ind = denD['leaves']
clean_axis(denAX)
cbSE, lookup = mapColors2Labels(labels, cmap=cmap, returnLookup=True)
axi = cbAX.imshow([[x] for x in cbSE.iloc[ind].values],
interpolation='nearest',
aspect='auto',
origin='lower')
clean_axis(cbAX)
colorLegend(list(lookup.values()), list(lookup.keys()), axh=denAX)
def update(self, t, alpha, beta, x_basis=False):
from matplotlib import pyplot as pl
ax = self.ax
pl.sca(ax)
pl.cla()
x_basis = self.x_basis
if x_basis:
from sglib import ip, col
from numpy import sqrt
alpha_x = ip(col(1,1)/sqrt(2), col(alpha,beta))
beta_x = ip(col(1,-1)/sqrt(2), col(alpha,beta))
alpha = alpha_x; beta = beta_x
prob_plus = alpha*alpha.conjugate()
prob_minus = beta*beta.conjugate()
pos_plus = self.pos_plus
pos_minus = self.pos_minus
width = self.width
p1 = pl.bar([pos_plus], prob_plus, width, color='blue')
p2 = pl.bar([pos_minus], prob_minus, width, color='red')
pl.xticks(self.xticks, self.xlabels)
pl.ylabel('Probability')
pl.yticks(self.yticks)
pl.xlim(0, 1.1)
pl.ylim(0, 1.1)
pl.draw()
pl.show()
def draw_all(self):
for i in range(0, len(self.cpustats)):
ax = plt.subplot(2, len(self.cpustats)/2, i)
plt.sca(ax)
self.draw(self.cpustats[i], plt)
plt.show()
def updateImage(self):
# clear plot
plt.sca(self.ax)
plt.cla()
self.initImage()
# Fetch data from SQL database
# try:
if self.current_plot == 'time_stack':
x,y = self.main_widget.imageAquisitionTab.stack.fetch_mologram_data()
elif self.current_plot == 'post_processing':
x,y = self.main_widget.postProcessingTab.get_time_and_molo_intensity()
else:
print('Current plot is empty')
return
# except Exception:
# print('Couldn\'t get data to plot')
# return
# plot appearance
plt.xlim([0, np.amax(x)])
plt.ylim([np.amin(y),np.amax(y)])
self.figure.tight_layout()
# add titles
self.ax.set_xlabel('Time (s)')
self.ax.set_ylabel('Mologram Intensity (arbitrary units)')
self.ax.ticklabel_format(style='sci', axis='y', scilimits=(0,0))
# plot data and show canvas
self.points = plt.plot(x,y)
self.canvas.draw()
def plot_confusion_matrix(cm, classes, ax,
normalize=False,
title='Confusion matrix',
cmap=plt.cm.Blues):
"""
This function prints and plots the confusion matrix.
Normalization can be applied by setting `normalize=True`.
"""
print(cm)
print('')
ax.imshow(cm, interpolation='nearest', cmap=cmap)
ax.set_title(title)
tick_marks = np.arange(len(classes))
plt.xticks(tick_marks, classes, rotation=45)
plt.sca(ax)
plt.yticks(tick_marks, classes)
fmt = '.2f' if normalize else 'd'
thresh = cm.max() / 2.
for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
ax.text(j, i, format(cm[i, j], fmt),
horizontalalignment="center",
color="white" if cm[i, j] > thresh else "black")
ax.set_ylabel('True label')
ax.set_xlabel('Predicted label')
def map(self, func, *args, **kwargs):
"""
Apply a plotting function to each facet's subset of the data.
Parameters
----------
func : callable
A plotting function that takes data and keyword arguments. It
must plot to the currently active matplotlib Axes and take a
`color` keyword argument. If faceting on the `hue` dimension,
it must also take a `label` keyword argument.
args : strings
Column names in self.data that identify variables with data to
plot. The data for each variable is passed to `func` in the
order the variables are specified in the call.
kwargs : keyword arguments
All keyword arguments are passed to the plotting function.
Returns
-------
self : FacetGrid object
"""
import matplotlib.pyplot as plt
for ax, namedict in zip(self.axes.flat, self.name_dicts.flat):
if namedict is not None:
data = self.data[namedict]
plt.sca(ax)
innerargs = [data[a].values for a in args]
func(*innerargs, **kwargs)
return self
def visualize_predictions(prediction_seqs, label_seqs, num_classes,
fig_width=6.5, fig_height_per_seq=0.5):
""" Visualize predictions vs. ground truth.
Args:
prediction_seqs: A list of int NumPy arrays, each with shape
`[duration, 1]`.
label_seqs: A list of int NumPy arrays, each with shape `[duration, 1]`.
num_classes: An integer.
fig_width: A float. Figure width (inches).
fig_height_per_seq: A float. Figure height per sequence (inches).
Returns:
A tuple of the created figure, axes.
"""
num_seqs = len(label_seqs)
max_seq_length = max([seq.shape[0] for seq in label_seqs])
figsize = (fig_width, num_seqs*fig_height_per_seq)
fig, axes = plt.subplots(nrows=num_seqs, ncols=1,
sharex=True, figsize=figsize)
for pred_seq, label_seq, ax in zip(prediction_seqs, label_seqs, axes):
plt.sca(ax)
plot_label_seq(label_seq, num_classes, 1)
plot_label_seq(pred_seq, num_classes, -1)
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
plt.xlim(0, max_seq_length)
plt.ylim(-2.75, 2.75)
plt.tight_layout()
return fig, axes
def visualize(model, filename):
values = model.all_params()
cols = OrderedDict()
for key, value in values.items():
if isinstance(value, np.ndarray): # TODO group by prefix, showing W and b side by side
cols.setdefault(key[:-2] + key[-1] if len(key) > 1 and key[-2] in "bW" else key, []).append((key, value))
_, axes = plt.subplots(2, len(cols), figsize=(5*len(cols), 10)) # TODO https://stackoverflow.com/a/13784887/223267
plt.tight_layout()
for j, col in enumerate(tqdm(cols.values(), unit="param", desc=filename)):
for i in range(2):
axis = axes[i, j] if len(values) > 1 else axes
if len(col) <= i:
plt.delaxes(axis)
else:
plt.sca(axis)
key, value = col[i]
plt.colorbar(plt.pcolormesh(smooth(value)))
for pattern, repl in REPLACEMENTS: # TODO map 0123->ifoc
key = re.sub(pattern, repl, key)
plt.title(key)
for axis in (plt.gca().xaxis, plt.gca().yaxis):
axis.set_major_locator(MaxNLocator(integer=True))
output_file = filename + ".png"
plt.savefig(output_file)
plt.clf()
print("Saved '%s'." % output_file)
def plotrfACC():
#data = json.loads(open('rf_accs.json').read())
data = json.loads(open('rf_accs_top3.json').read())
data = json.loads(open('rf_accs_nowindow.json').read())
nLetter = 3 #14
data["texts/ADHD_various_half/"] = [data["texts/ADHD_various_half/"][i] for i in [1,2,3]]
sns.set_style("dark")
#f, (ax1, ax2) = plt.subplots(1, 2)
f, ax1 = plt.subplots()
bar1 = ax1.bar(range(nLetter),data["texts/ADHD_various_half/"])
ax1.set_title('RF accs for half SAX')
plt.sca(ax1)
plt.xticks(np.arange(nLetter) + .4, range(3,nLetter+3))
plt.xlabel('# of bins (letters)/word')
ax1.set_ylim([0.6,0.9])
#bar2 = ax2.bar(range(nLetter),data["texts/ADHD_various_full/"])
#ax2.set_title('RF accs for full SAX')
#plt.sca(ax2)
#plt.xticks(np.arange(nLetter) + .4, range(2,nLetter+2))
#plt.xlabel('# of bins (letters)/word')
#ax2.set_ylim([0.6,0.9])
plt.show()
plt.xlabel("covariate", fontsize=15)
plt.xticks(np.arange(36), covariates, fontsize=12, rotation=90)
plt.axhline(y=0, color="k", alpha=0.5, ls="--")
plt.tight_layout()
# Get expected states:
posterior_probs = [
bandit_glmhmm.expected_states(data=meta_true_choices, input=meta_inpts)[0]
for meta_true_choices, meta_inpts in zip(meta_true_choices, meta_inpts)
]
fig, ax = plt.subplots(5, 4, figsize=(15, 20))
for i in range(num_sessions):
column = int(np.floor(i / 5))
row = i % 5
plt.sca(ax[row, column])
for k in range(num_states):
plt.plot(posterior_probs[i][:, k],
label="State " + str(k + 1),
lw=2,
color=cols[k])
plt.ylim((-0.01, 1.01))
plt.yticks([0, 0.5, 1], fontsize=10)
plt.xlabel("trial #", fontsize=15)
plt.ylabel("p(state)", fontsize=15)
plt.tight_layout()
plt.plot(log_likelihoods.mean(axis=1))
plt.xticks(np.arange(5), np.arange(1, 6))
plt.xlabel('Number of latents')
plt.ylabel('LL')
0] = 1e-8 #replace zero values with a small value to avoid dividing by zero.
KL[i] = stats.entropy(int_density[:, i], qk=int_density_prior[:, i])
ax3_twin = ax3.twinx()
ax3_twin.plot(x_density[1, :], KL, 'r')
ax3_twin.set_ylabel('KL divergence', fontsize=16, color='r')
ax3_twin.tick_params('y', colors='r', labelsize=16)
ax3_twin.tick_params(direction='out', pad=5)
# produce an image of the prior intensity:
threshold_prior = 0.001
int_density_prior_refined = int_density_prior.copy()
int_density_prior_refined[
int_density_prior_refined < threshold_prior] = 0.0
fig8, ax8 = plt.subplots(figsize=(8, 5))
plt.sca(ax8)
plt.imshow(int_density_prior_refined,
origin='lower',
cmap=cm.jet,
extent=(x_density_prior[0, 0], x_density_prior[0, -1],
y_density_prior[0, 0], y_density_prior[-1, 0]),
aspect='auto',
norm=LogNorm(vmin=0.001, vmax=0.1),
interpolation="nearest")
ax8.set_xlim(x_density_prior[0, 0], x_density_prior[0, -1])
ax8.set_ylim(y_density_prior[0, 0], y_density_prior[-1, 0])
ax8.set_xlabel('Age/yr', fontsize=16)
ax8.set_ylabel('Intensity/$\mu$T', fontsize=16)
cb = plt.colorbar(ticks=[
0.001, 0.002, 0.003, 0.004, 0.005, 0.006, 0.007, 0.008, 0.009, 0.01
df_emp1 = df_emp.groupby(['rango_n_transac']).size().to_frame()
df_emp1.rename(columns={0: 'donante empresas'}, inplace=True)
for frame in [df_co1, df_int1, df_ext1, df_emp1]:
frame.plot(ax=ax1)
ax1.set_xlabel('Número de transacciones promedio al año')
ax1.set_ylabel('Cantidad de donantes')
ax1.set_title("Número promedio de transacciones al año de donantes")
df_ext1 = df_ext.groupby(['rango_n_transac'
])['donacion_total'].sum().to_frame()
df_ext1.rename(columns={'donacion_total': 'donante particulares externas'},
inplace=True)
# print(df_ext1)
df_int1 = df_int.groupby(['rango_n_transac'
])['donacion_total'].sum().to_frame()
df_int1.rename(columns={'donacion_total': 'donante particulares internas'},
inplace=True)
df_co1 = df_co.groupby(['rango_n_transac'])['donacion_total'].sum().to_frame()
df_co1.rename(columns={'donacion_total': 'donante cooperadoras'}, inplace=True)
df_emp1 = df_emp.groupby(['rango_n_transac'
])['donacion_total'].sum().to_frame()
df_emp1.rename(columns={'donacion_total': 'donante empresas'}, inplace=True)
for frame in [df_co1, df_int1, df_ext1, df_emp1]:
frame.plot(ax=ax2)
ax2.set_ylabel('Monto acumulado (en miles de millones)')
ax2.set_xlabel('Número de transacciones promedio al año')
ax2.set_title("Monto total acumulado por transacciones")
for ax in fig.axes:
plt.sca(ax)
plt.xticks(rotation=0)
fig.savefig('./images/todos/descriptivo_14.png')
]
axBW.set_xticks(range(2))
axBW.set_xlim([-0.5, 1.5])
extraplots.boxoff(axBW)
extraplots.set_ticks_fontsize(axBW, fontSizeTicks)
axBW.set_xticklabels(tickLabels, fontsize=fontSizeLabels, rotation=45)
zstat, pVal = stats.mannwhitneyu(nD1PopStat,
D1PopStat) #Nick used stats.ranksum
messages.append("{} p={}".format(popStatCol, pVal))
yDataMax = max([max(D1PopStat), max(nD1PopStat)])
yStars = yDataMax + yDataMax * starYfactor
yStarHeight = (yDataMax * starYfactor) * starHeightFactor
plt.sca(axBW)
starString = None if pVal < 0.05 else 'n.s.'
extraplots.significance_stars([0, 1],
yStars,
yStarHeight,
starMarker='*',
starSize=fontSizeStars + 2,
starString=starString,
gapFactor=starGapFactor)
plt.hold(1)
#
plt.hold(True)
if PANELS[8]:
popStatCol = 'thresholdFRA'
nD1PopStat = D1.query('{} == {}'.format(popStatCol,
def perform_fft(input_info, plot=False):
print("\nspawn FFT process nr : ", input_info["process_nr"])
existing_shm = shared_memory.SharedMemory(name=input_info["shm"].name)
# check if poly calculation is on to be executed, because then there must be an additional key in the dictionary
if "poly" in input_info.keys():
poly = True
reference_to_data_block = numpy.ndarray(
input_info["dim"], dtype=numpy.float64,
buffer=existing_shm.buf)[:, input_info["from"]:input_info["to"], :]
else:
poly = False
reference_to_data_block = numpy.ndarray(
input_info["dim"], dtype=numpy.float64,
buffer=existing_shm.buf)[:, input_info["from"]:input_info["to"], :]
# get data to process out of buffer
print("ref data dtype: ", reference_to_data_block.dtype)
print("\n")
# reshape data from buffer to 2d matrix with the time as y coords and x as the values
data_mat = reference_to_data_block.reshape(
reference_to_data_block.shape[0],
reference_to_data_block.shape[1] * reference_to_data_block.shape[2])
# strore orig time, cols and row information - needed for reshaping
orig_time = reference_to_data_block.shape[0]
orig_rows = reference_to_data_block.shape[1]
orig_cols = reference_to_data_block.shape[2]
# if plots are whished
if plot:
existing_shm_qual = shared_memory.SharedMemory(
name=input_info["shm_qual"].name)
if poly:
reference_to_qual_block = numpy.ndarray(
input_info["dim"],
dtype=numpy.float64,
buffer=existing_shm_qual.buf
)[:, input_info["from"]:input_info["to"], :]
else:
reference_to_qual_block = numpy.ndarray(
input_info["dim"],
dtype=numpy.float64,
buffer=existing_shm_qual.buf
)[:, input_info["from"]:input_info["to"], :]
qual_weights = input_info["weights"]
qual_factor = 1
qual_mat = reference_to_qual_block.reshape(
reference_to_qual_block.shape[0],
reference_to_qual_block.shape[1] *
reference_to_qual_block.shape[2])
qual_mat[qual_mat == 255] = numpy.nan
print("Data Mat Shape", data_mat.shape)
# setting int Nan value to numpy.nan --> transforms dataytpe to floa64!!!
data_mat = numpy.where(data_mat == 32767, numpy.nan, data_mat)
n = data_mat.shape[0]
t = numpy.arange(0, n, 1)
# iter through
for i in range(0, data_mat.shape[1], 1):
data_mat_v_nan = numpy.isfinite(data_mat[:, i])
data_mat_v_t = numpy.arange(0, len(data_mat_v_nan), 1)
if False in data_mat_v_nan:
try:
# interpolate on that spots
data_mat_v_interp = numpy.round(
numpy.interp(data_mat_v_t, data_mat_v_t[data_mat_v_nan],
data_mat[:, i][data_mat_v_nan]))
# calculate the fft
f_hat = numpy.fft.fft(data_mat_v_interp, n)
# and the power spectrum - which frequencies are dominant
power_spectrum = f_hat * numpy.conj(f_hat) / n
# get the max power value
max_fft_spectr_value = numpy.max(power_spectrum)
# set it to zeros so one can find those frequencies that are far lower and important but still no noise
power_spec_no_max = numpy.where(
power_spectrum == max_fft_spectr_value, 0, power_spectrum)
threshold_remaining_values = numpy.nanmax(
power_spec_no_max) / 2
indices = power_spectrum > threshold_remaining_values
f_hat = indices * f_hat
ffilt = numpy.fft.ifft(f_hat)
if plot:
if i <= 3:
print("proces nr %d i == %d" %
(input_info["process_nr"], i))
print("data mat: ", data_mat[:, i])
print("data_mat_v_interp", data_mat_v_interp)
# print("data mat:", data_mat[:, i])
# print("data_mat.dtype: ", data_mat.dtype)
# print("data_mat_interp.dtype: ", data_mat_v_interp.dtype)
ffilt = numpy.round(ffilt).astype(numpy.int16)
# print("\ntransfrom to int16: ", data_mat_v_interp)
# print("\ndata_mat_interp.dtype: ", data_mat_v_interp.dtype)
fig, axs = plt.subplots(3, 1)
good_qual = numpy.where(
qual_mat[:, i] == qual_weights[0], qual_weights[0],
numpy.nan) * qual_factor
okay_qual = numpy.where(
qual_mat[:, i] == qual_weights[1], qual_weights[1],
numpy.nan) * qual_factor
bad_qual = numpy.where(
qual_mat[:, i] == qual_weights[2], qual_weights[2],
numpy.nan) * qual_factor
really_bad_qual = numpy.where(
qual_mat[:, i] == qual_weights[3], qual_weights[3],
numpy.nan) * qual_factor
nan_values = numpy.where(qual_mat[:, i] == 255,
qual_weights[3], numpy.nan)
plt.sca(axs[0])
plt.plot(t,
data_mat[:, i],
color='c',
LineWidth=3,
label="raw data")
plt.plot(t,
data_mat_v_interp,
color='k',
LineWidth=1,
linestyle='--',
label='lin interp')
plt.plot(t,
ffilt,
color="k",
LineWidth=2,
label='FFT Filtered')
plt.plot(t, good_qual, 'go', label="Good Quality")
plt.plot(t, okay_qual, 'yo', label="Okay Quality")
plt.plot(t,
bad_qual,
'o',
color='orange',
label="Bad Quality")
plt.plot(t,
really_bad_qual,
'ro',
label="Really Bad Quality")
plt.plot(t, nan_values, 'ko', label="NaN Values")
plt.xlim(t[0], t[-1])
plt.ylabel("Intensity [%]")
plt.xlabel("Time [days]")
plt.legend()
plt.sca(axs[1])
plt.plot(t,
power_spectrum,
color="c",
LineWidth=2,
label="Noisy")
plt.plot(t,
power_spectrum,
'b*',
LineWidth=2,
label="Noisy")
plt.plot(t[0], t[-1])
plt.xlabel("Power Spectrum [Hz]")
plt.ylabel("Power")
plt.title(
"Power Spectrum Analyses - Max: {} - Threshold: {}"
.format(max_fft_spectr_value,
numpy.nanmean(power_spectrum)))
plt.sca(axs[2])
plt.plot(t,
power_spec_no_max,
color="c",
LineWidth=2,
label="Noisy")
plt.plot(t,
power_spec_no_max,
'b*',
LineWidth=2,
label="Noisy")
plt.plot(t[0], t[-1])
plt.xlabel("Power Spectrum no max [Hz]")
plt.ylabel("Power")
plt.title(
"Power Spectrum Analysis - removed big max {} - Max: {} - Threshold: {}"
.format(max_fft_spectr_value,
numpy.nanmax(power_spec_no_max),
threshold_remaining_values))
# plot data
plt.show()
data_mat[:, i] = ffilt
except:
# gets triggered most if there are only nans in the array
continue
else:
# calculate the fft
f_hat = numpy.fft.fft(data_mat[:, i], n)
# and the power spectrum - which frequencies are dominant
power_spectrum = f_hat * numpy.conj(f_hat) / n
# get the max power value
max_fft_spectr_value = numpy.max(power_spectrum)
# set it to zeros so one can find those frequencies that are far lower and important but still no noise
power_spec_no_max = numpy.where(
power_spectrum == max_fft_spectr_value, 0, power_spectrum)
threshold_remaining_values = numpy.nanmax(power_spec_no_max) / 2
indices = power_spectrum > threshold_remaining_values
f_hat = indices * f_hat
ffilt = numpy.fft.ifft(f_hat)
data_mat[:, i] = ffilt
# transorm float64 back to INT16!!
# save interpolation results on the shared memory object
if poly:
reference_to_data_block[:] = numpy.round(
data_mat.reshape(orig_time, orig_rows, orig_cols))
else:
reference_to_data_block[:] = numpy.round(
data_mat.reshape(orig_time, orig_rows, orig_cols))
ax[0].set_xlim(-2, 2)
ax[0].set_ylim(-1, 1)
ax[1].set_xlim(0, max(t))
ax[1].set_ylim(min(x) * 1.2, max(x * 1.2))
ax[1].set_ylabel('X')
ax[2].set_xlim(0, max(t))
ax[2].set_ylim(min(x) * 1.2, max(x * 1.2))
ax[2].set_ylabel('Y')
line, = ax[0].plot([], [], lw=2, alpha=0.5)
dot, = ax[0].plot([], [], 'ro', ms=5)
marker1, = ax[0].plot([], [], 'k', alpha=0.5, lw=2)
marker2, = ax[1].plot([], [], 'k', alpha=0.5, lw=2)
var1, = ax[1].plot([], [], lw=2, alpha=0.5)
var2, = ax[2].plot([], [], lw=2, alpha=0.5)
for this_ax in ax:
plt.sca(this_ax)
plt.axis('off')
# initialization function: plot the background of each frame
def init():
line.set_data([], [])
dot.set_data([], [])
marker1.set_data([], [])
marker2.set_data([], [])
var1.set_data([], [])
var2.set_data([], [])
return dot, line, var1, var2, marker1, marker2
# animation function. This is called sequentially
def param_plot(x, func, params, **kwargs):
"""
Plot a function with user-adjustable parameters.
Parameters
----------
x : array_like
Independent (x-axis) variable.
func : function
Function that takes as its first argument an independent variable and
as subsequent arguments takes parameters. It should return an output
array the same dimension as `x`, which is plotted as the y-variable.
params : dict
Dictionary whose keys are strings named exactly as the parameter
arguments to `func` are. [More info on options]
Returns
-------
final_params : dict
A dict whose keys are the same as `params` and whose values correspond
to the values selected by the slider. `final_params` will continue to
change until the figure is closed, at which point it has the final
parameter values the user chose. This is useful for hand-fitting
curves.
"""
params = _initialize_params(params)
flat_params = {k: _unitify(v, v['init']) for k, v in params.iteritems()}
# Set up figure and axes that we'll plot on
fig = plt.figure()
ax0 = fig.add_axes([0, 0, 1, 1])
l, = ax0.plot(x, func(x, **flat_params), **kwargs)
# Function for redrawing curve when a parameter value is changed
def update(val):
param_args = {
k: _unitify(v, v['slider'].val)
for k, v in params.iteritems()
}
l.set_ydata(func(x, **param_args))
fig.canvas.draw_idle()
# Update values in flat_params so end user can get their final values
flat_params.update(param_args)
# Create axes and sliders for each parameter
for i, (name, vals) in enumerate(params.iteritems()):
ax = plt.axes(_slider_bbox(i).bounds)
slider = Slider(ax, vals['label'], vals['min'], vals['max'],
vals['init'])
slider.on_changed(update)
vals['ax'] = ax
vals['slider'] = slider
# Function for auto-scaling sliders to (ironically) keep them fixed
def resize_func(event):
ax0.set_position(_main_ax_bbox(ax0, len(params), fig).bounds)
for i, val in enumerate(params.itervalues()):
box = _slider_bbox(i)
val['ax'].set_position(box)
fig.canvas.mpl_connect('resize_event', resize_func)
# Set current axes to the one you'd expect
plt.sca(ax0)
return flat_params
def state_vars_timeseries(rec,
modelspec,
ax=None,
state_colors=None,
decimate_by=1,
channel=None):
if ax is not None:
plt.sca(ax)
pred = rec['pred']
resp = rec['resp']
fs = resp.fs
chanidx = get_channel_number(resp, channel)
r1 = resp.as_continuous()[chanidx, :].T * fs
p1 = pred.as_continuous()[chanidx, :].T * fs
nnidx = np.isfinite(p1)
r1 = r1[nnidx]
p1 = p1[nnidx]
if decimate_by > 1:
r1 = scipy.signal.decimate(r1, q=decimate_by, axis=0)
p1 = scipy.signal.decimate(p1, q=decimate_by, axis=0)
fs /= decimate_by
t = np.arange(len(r1)) / fs
plt.plot(t, r1, linewidth=1, color='gray')
plt.plot(t, p1, linewidth=1, color='black')
mmax = np.nanmax(p1) * 0.8
if 'state' in rec.signals.keys():
s = None
g = None
d = None
for m in modelspec:
if 'state_dc_gain' in m['fn']:
g = np.array(m['phi']['g'])
d = np.array(m['phi']['d'])
if len(g) < 10:
s = ",".join(rec["state"].chans)
g_string = np.array2string(g, precision=3)
d_string = np.array2string(d, precision=3)
s += " g={} d={} ".format(g_string, d_string)
else:
s = None
num_vars = rec['state'].shape[0]
ts = rec['state'].as_continuous().copy()
if state_colors is None:
state_colors = [None] * num_vars
print(nnidx.shape)
print(ts.shape)
for i in range(1, num_vars):
st = ts[i, :].T
if decimate_by > 1:
st = scipy.signal.decimate(st[nnidx], q=decimate_by, axis=0)
else:
st = st[nnidx]
st = st / np.nanmax(st) * mmax - (0.1 + i) * mmax
plt.plot(t, st, linewidth=1, color=state_colors[i - 1])
if g is not None:
if g.ndim == 1:
tstr = "{} (d={:.3f},g={:.3f})".format(
rec['state'].chans[i], d[i], g[i])
else:
tstr = "{} (d={:.3f},g={:.3f})".format(
rec['state'].chans[i], d[0, i], g[0, i])
else:
tstr = "{}".format(rec['state'].chans[i])
plt.text(t[0], (-i + 0.1) * mmax, tstr)
ax = plt.gca()
# plt.text(0.5, 0.9, s, transform=ax.transAxes,
# horizontalalignment='center')
# if s:
# plt.title(s, fontsize=8)
plt.xlabel('time (s)')
plt.axis('tight')
ax_remove_box(ax)
def test_rupture_depth(interactive=False):
DIP = 17.0
WIDTH = 20.0
GRIDRES = 0.1
names = [
'single', 'double', 'triple', 'concave', 'concave_simple', 'ANrvSA'
]
means = [
3.1554422780092461, 2.9224454569459781, 3.0381968625073563,
2.0522694624400271, 2.4805390352818755, 2.8740121776209673
]
stds = [
2.1895293825074575, 2.0506459673526174, 2.0244588429154402,
2.0112565876976416, 2.1599789955270019, 1.6156220309120068
]
xp0list = [
np.array([118.3]),
np.array([10.1, 10.1]),
np.array([10.1, 10.1, 10.3]),
np.array([10.9, 10.5, 10.9]),
np.array([10.9, 10.6]),
np.array([
-76.483, -76.626, -76.757, -76.99, -77.024, -76.925, -76.65,
-76.321, -75.997, -75.958
])
]
xp1list = [
np.array([118.3]),
np.array([10.1, 10.3]),
np.array([10.1, 10.3, 10.1]),
np.array([10.5, 10.9, 11.3]),
np.array([10.6, 10.9]),
np.array([
-76.626, -76.757, -76.99, -77.024, -76.925, -76.65, -76.321,
-75.997, -75.958, -76.006
])
]
yp0list = [
np.array([34.2]),
np.array([34.2, 34.5]),
np.array([34.2, 34.5, 34.8]),
np.array([34.2, 34.5, 34.8]),
np.array([35.1, 35.2]),
np.array([
-52.068, -51.377, -50.729, -49.845, -49.192, -48.507, -47.875,
-47.478, -47.08, -46.422
])
]
yp1list = [
np.array([34.5]),
np.array([34.5, 34.8]),
np.array([34.5, 34.8, 35.1]),
np.array([34.5, 34.8, 34.6]),
np.array([35.2, 35.4]),
np.array([
-51.377, -50.729, -49.845, -49.192, -48.507, -47.875, -47.478,
-47.08, -46.422, -45.659
])
]
for i in range(0, len(xp0list)):
xp0 = xp0list[i]
xp1 = xp1list[i]
yp0 = yp0list[i]
yp1 = yp1list[i]
name = names[i]
mean_value = means[i]
std_value = stds[i]
zp = np.zeros(xp0.shape)
strike = azimuth(xp0[0], yp0[0], xp1[-1], yp1[-1])
widths = np.ones(xp0.shape) * WIDTH
dips = np.ones(xp0.shape) * DIP
strike = [strike]
origin = Origin({
'id': 'test',
'lon': 0,
'lat': 0,
'depth': 5.0,
'mag': 7.0,
'netid': 'us',
'network': '',
'locstring': '',
'time': HistoricTime.utcfromtimestamp(time.time())
})
rupture = QuadRupture.fromTrace(xp0,
yp0,
xp1,
yp1,
zp,
widths,
dips,
origin,
strike=strike)
# make a grid of points over both quads, ask for depths
ymin = np.nanmin(rupture.lats)
ymax = np.nanmax(rupture.lats)
xmin = np.nanmin(rupture.lons)
xmax = np.nanmax(rupture.lons)
xmin = np.floor(xmin * (1 / GRIDRES)) / (1 / GRIDRES)
xmax = np.ceil(xmax * (1 / GRIDRES)) / (1 / GRIDRES)
ymin = np.floor(ymin * (1 / GRIDRES)) / (1 / GRIDRES)
ymax = np.ceil(ymax * (1 / GRIDRES)) / (1 / GRIDRES)
geodict = GeoDict.createDictFromBox(xmin, xmax, ymin, ymax, GRIDRES,
GRIDRES)
nx = geodict.nx
ny = geodict.ny
depths = np.zeros((ny, nx))
for row in range(0, ny):
for col in range(0, nx):
lat, lon = geodict.getLatLon(row, col)
depth = rupture.getDepthAtPoint(lat, lon)
depths[row, col] = depth
np.testing.assert_almost_equal(np.nanmean(depths), mean_value)
np.testing.assert_almost_equal(np.nanstd(depths), std_value)
if interactive:
fig, axes = plt.subplots(nrows=2, ncols=1)
ax1, ax2 = axes
xdata = np.append(xp0, xp1[-1])
ydata = np.append(yp0, yp1[-1])
plt.sca(ax1)
plt.plot(xdata, ydata, 'b')
plt.sca(ax2)
im = plt.imshow(depths, cmap='viridis_r') # noqa
ch = plt.colorbar() # noqa
fname = os.path.join(os.path.expanduser('~'),
'quad_%s_test.png' % name)
print('Saving image for %s quad test... %s' % (name, fname))
plt.savefig(fname)
plt.close()
def state_var_psth_from_epoch(rec,
epoch,
psth_name='resp',
psth_name2='pred',
state_sig='pupil',
ax=None,
colors=None,
channel=None,
decimate_by=1):
"""
Plot PSTH averaged across all occurences of epoch, grouped by
above- and below-average values of a state signal (state_sig)
"""
# TODO: Does using epochs make sense for these?
if ax is not None:
plt.sca(ax)
chanidx = get_channel_number(rec[psth_name], channel)
fs = rec[psth_name].fs
d = rec[psth_name].get_epoch_bounds('PreStimSilence')
PreStimSilence = np.mean(np.diff(d)) - 0.5 / fs
d = rec[psth_name].get_epoch_bounds('PostStimSilence')
if d.size > 0:
PostStimSilence = np.min(np.diff(d)) - 0.5 / fs
dd = np.diff(d)
dd = dd[dd > 0]
else:
dd = np.array([])
if dd.size > 0:
PostStimSilence = np.min(dd) - 0.5 / fs
else:
PostStimSilence = 0
full_psth = rec[psth_name]
channel = 0
folded_psth = full_psth.extract_epoch(epoch)[:, [chanidx], :] * fs
if psth_name2 is not None:
full_psth2 = rec[psth_name2]
folded_psth2 = full_psth2.extract_epoch(epoch)[:, [chanidx], :] * fs
if state_sig == "each_passive":
raise ValueError("each_passive state not supported")
# extract high (=1) epochs from each passive state
else:
full_var = rec['state'].loc[state_sig]
folded_var = full_var.extract_epoch(epoch)
# remove masked out occurences if mask signal exists
if 'mask' in rec.signals.keys():
folded_mask = rec['mask'].extract_epoch(epoch)
keep_occurences = folded_mask[:, 0, 0]
folded_psth = folded_psth[keep_occurences, :, :]
folded_psth2 = folded_psth2[keep_occurences, :, :]
folded_var = folded_var[keep_occurences, :, :]
# print(state_sig)
# print(folded_var.shape)
# print(folded_mask.shape)
# print(np.sum(np.isfinite(folded_mask)))
if decimate_by > 1:
folded_psth = scipy.signal.decimate(folded_psth, q=decimate_by, axis=2)
folded_psth2 = scipy.signal.decimate(folded_psth2,
q=decimate_by,
axis=2)
fs /= decimate_by
# compute the mean state for each occurrence
m = np.nanmean(folded_var[:, 0, :], axis=1)
# compute the mean state across all occurrences
mean = np.nanmean(m)
gtidx = (m >= mean)
ltidx = np.logical_not(gtidx)
# low = response on epochs when state less than mean
if np.sum(ltidx):
low = np.nanmean(folded_psth[ltidx, :, :], axis=0).T
low2 = np.nanmean(folded_psth2[ltidx, :, :], axis=0).T
else:
low = np.ones(folded_psth[0, :, :].shape).T * np.nan
low2 = np.ones(folded_psth2[0, :, :].shape).T * np.nan
# high = response on epochs when state greater than or equal to mean
if np.sum(gtidx):
high = np.nanmean(folded_psth[gtidx, :, :], axis=0).T
high2 = np.nanmean(folded_psth2[gtidx, :, :], axis=0).T
else:
high = np.ones(folded_psth[0, :, :].shape).T * np.nan
high2 = np.ones(folded_psth2[0, :, :].shape).T * np.nan
title = state_sig
hv = np.nanmean(m[m >= mean])
if np.sum(m < mean) > 0:
lv = np.nanmean(m[m < mean])
if (hv > 0.95) and (hv < 1.05) and (lv > -0.05) and (lv < 0.05):
legend = ('Lo', 'Hi')
else:
legend = ('< Mean', '>= Mean')
timeseries_from_vectors([low, high],
fs=fs,
title=title,
ax=ax,
legend=legend,
time_offset=PreStimSilence,
colors=colors,
ylabel="sp/sec")
timeseries_from_vectors([low2, high2],
fs=fs,
title=title,
ax=ax,
linestyle='--',
time_offset=PreStimSilence,
colors=colors,
ylabel="sp/sec")
else:
timeseries_from_vectors([low, high],
fs=fs,
title=title,
ax=ax,
time_offset=PreStimSilence,
colors=colors,
ylabel="sp/sec")
timeseries_from_vectors([low2, high2],
fs=fs,
title=title,
ax=ax,
linestyle='--',
time_offset=PreStimSilence,
colors=colors,
ylabel="sp/sec")
ylim = ax.get_ylim()
xlim = ax.get_xlim()
ax.plot(np.array([0, 0]), ylim, 'k--')
ax.plot(np.array([xlim[1], xlim[1]]) - PostStimSilence, ylim, 'k--')
if state_sig == 'baseline':
ax.set_xlabel(epoch)
def generate_plot(partial_quota, number_of_instances, f, a0, a1, tmp_y2, tmp_x,
tmp_y, blocks_x, start_point, Xtics, yearly_quota, x_start,
finished, User_t, System_t, y_start2, y_end2, beginning_dt,
nutzergraph, fig, x_end, Data, filter_n, months):
if filter_n:
f.suptitle(str(filter_n), fontweight="bold")
else:
f.suptitle(str(Data[0]['Account'])[2:-1], fontweight="bold")
global daily_eff_days
global daily_eff_eff
fmt = "%Y-%m-%d-%H-%M" # standard format for Dates, year month, day, hour, minute
myFmt = mdates.DateFormatter('%b %y')
nothing = mdates.DateFormatter(' ')
monthly_cputime = []
monthly_used = []
effarray = []
tmp_y2.append(tmp_y[-1])
if partial_quota: #### drawing of the quota####
for i in range(0, number_of_instances
): # not possible for the last area, hence skipping it.
col = colorization(
D_.find_y_from_given_time(
datetime.datetime.fromtimestamp(blocks_x[i * 2 + 1]),
tmp_x, tmp_y) - D_.find_y_from_given_time(
datetime.datetime.fromtimestamp(blocks_x[i * 2]),
tmp_x, tmp_y), partial_quota)
coordinates_x = (datetime.datetime.fromtimestamp(blocks_x[i * 2]),
datetime.datetime.fromtimestamp(blocks_x[i * 2]),
datetime.datetime.fromtimestamp(blocks_x[i * 2 +
1]))
coordinates_y = [
tmp_y2[i * 2], tmp_y2[i * 2 + 1], tmp_y2[i * 2 + 1]
]
a0.fill_between(coordinates_x,
0,
coordinates_y,
color=col,
alpha=0.99)
monthly_cputime.append(tmp_y2[i * 2 + 1] - tmp_y2[i * 2])
value1 = D_.find_y_from_given_time(
datetime.datetime.fromtimestamp(blocks_x[-1]), tmp_x, tmp_y)
value2 = D_.find_y_from_given_time(
datetime.datetime.fromtimestamp(blocks_x[-2]), tmp_x, tmp_y)
col = colorization(value1 - value2, partial_quota)
coordinates_x = (datetime.datetime.fromtimestamp(blocks_x[-2]),
datetime.datetime.fromtimestamp(blocks_x[-2]),
datetime.datetime.fromtimestamp(blocks_x[-1]))
coordinates_y = (value2, value2 + partial_quota,
value2 + partial_quota)
a0.fill_between(coordinates_x, 0, coordinates_y, color=col, alpha=0.99)
# determines the last interval's color and draws it (uses the highest
# recorded value as the end value of the ongoing time span).
axis = plt.gca() # for plotting/saving the plot as it's own image
# Sets the visual borders for the graphs; area of occurring values (main graph) +- 5%.
if start_point: # setting the beginning and end of the graph
beginning = start_point.timestamp()
end = start_point.timestamp() + 365 * 24 * 3600
beginning = beginning - 30 * 24 * 3600
end = end + 30 * 24 * 3600
extrapolation_x = []
extrapolation_y = []
if len(tmp_y2) < 3:
tmp_y2.append(0)
tmp_y2.append(0)
usedmonths = 0
usedmonths += 12 * (int(x_end.strftime("%Y")) -
int(x_start.strftime("%Y")))
usedmonths += int(x_end.strftime("%m")) - int(x_start.strftime("%m"))
monthsleft = int(months - usedmonths)
if yearly_quota and len(tmp_x) >= 1:
extrapolation_point_x = D_.first_of_month(x_end)
extrapolation_point_y = D_.find_y_from_given_time(
tmp_x[-1], tmp_x, tmp_y)
extrapolation_point_y = max(
extrapolation_point_y,
D_.find_y_from_given_time(D_.first_of_month(extrapolation_point_x),
tmp_x, tmp_y) + partial_quota)
extrapolation_x.append(D_.first_of_month(extrapolation_point_x))
extrapolation_x.append(D_.first_of_month(extrapolation_point_x))
extrapolation_x.append(
D_.first_of_month(
datetime.datetime.fromtimestamp(
extrapolation_point_x.timestamp() + 2851200)))
extrapolation_y.append(
D_.find_y_from_given_time(extrapolation_point_x, tmp_x, tmp_y))
extrapolation_y.append(
max(extrapolation_y[0] + partial_quota, tmp_y[-1]))
extrapolation_y.append(extrapolation_y[-1])
expoint_y = extrapolation_y[-1]
extrapolation_y[-2] = tmp_y[-1]
extrapolation_y[-1] = tmp_y[-1]
expoint_y = extrapolation_y[-1]
xtr_pt_x = extrapolation_point_x
xtr_pt_y = extrapolation_point_y
for i in range(1,
monthsleft): # The three points required for each block
extrapolation_x.append(
D_.first_of_month(
datetime.datetime.fromtimestamp(xtr_pt_x.timestamp() +
i * 2851200)))
extrapolation_x.append(
D_.first_of_month(
datetime.datetime.fromtimestamp(xtr_pt_x.timestamp() +
i * 2851200)))
extrapolation_x.append(
D_.first_of_month(
datetime.datetime.fromtimestamp(xtr_pt_x.timestamp() +
(i + 1) * 2851200)))
extrapolation_y.append(expoint_y + (i - 1) * partial_quota)
extrapolation_y.append(expoint_y + i * partial_quota)
extrapolation_y.append(expoint_y + i * partial_quota)
if monthsleft:
a0.plot(extrapolation_x[3:], extrapolation_y[3:], "black")
if monthsleft:
extrapolation_y.append(0)
else:
extrapolation_y = [0]
beg_14_months = beginning + 36817200
fourteen_dt = datetime.datetime.fromtimestamp(beg_14_months)
# Print statements, to give feedback either onscreen or into a dedicated file to be piped into.
print('The accumulated TotalCPU time is',
int((User_t[-1] + System_t[-1]) * 100) / 100, "hours")
print('and the number of accumulated corehours is',
int(tmp_y[-1] * 100) / 100)
efficiency = (User_t[-1] + System_t[-1]) / tmp_y[-1]
# Added rounding to the efficiency percentage feedback.
print('Which results in an efficiency of',
int(efficiency * 10000) / 100 + 0.005, "%")
if efficiency < 0 or efficiency > 1:
print(
"Efficiency is outside of it's boundaries, valid is only between 0 and 1"
)
accum_total_time = np.zeros(len(tmp_x))
for i in range(0, len(accum_total_time)):
accum_total_time[i] = User_t[i] + System_t[i]
delta = [0]
total_time = []
total_time.append(accum_total_time[0])
total_time.append(accum_total_time[0])
difference = [0]
for i in range(1, len(accum_total_time)):
total_time.append(accum_total_time[i] - accum_total_time[i - 1])
delta.append(100 * ((accum_total_time[i] - accum_total_time[i - 1]) /
(tmp_y[i] - tmp_y[i - 1])))
if delta[i] > 100:
a = 0
if yearly_quota: # ensuring that the extrapolated quota is still in frame
a0.set_ylim([
y_start2 - (0.05 * y_end2),
max(tmp_y[-1], max(extrapolation_y), max(coordinates_y)) * 1.2
])
# print("limit",a0.get_ylim()[1])
else: # No quota given, image is focused around occupied and utilized resources.
print("NO YEARLY DETECTED")
a0.set_ylim([y_start2 - (0.05 * y_end2), tmp_y[-1] * 1.05])
##### Creation of patches for Legend #####
red_patch = mpatches.Patch(color='#ff0000', alpha=0.7, label='>=150%')
orange_patch = mpatches.Patch(color='#ffa500',
alpha=0.7,
label='>=110%,<150%')
green_patch = mpatches.Patch(color='#008000',
alpha=0.8,
label='>=70%,<110%')
light_green_patch = mpatches.Patch(color='#81c478',
alpha=0.8,
label='<70%')
grey_patch = mpatches.Patch(color='dimgrey',
alpha=0.75,
label='Allocated corehours')
yellow_patch = mpatches.Patch(color='#d9e72e',
alpha=0.49,
label='Utilized corehours')
black_patch = mpatches.Patch(color='black',
alpha=1,
label='Granted corehours')
a0.plot(tmp_x, accum_total_time, '#d9e72e') # plotting the TotatlCPU Graph
if yearly_quota: # Legends for if there is a quota, or a shorter Legend in case there isn't.
a0.legend(handles=[
red_patch, orange_patch, green_patch, light_green_patch,
grey_patch, yellow_patch, black_patch
])
else:
a0.legend(handles=[grey_patch, yellow_patch])
a0.fill_between(tmp_x, 0, accum_total_time, color='#d9e72e',
alpha=0.70) # plotting the area below TotalCPU graph
a0.plot(tmp_x, tmp_y, 'dimgrey', fillstyle='bottom',
alpha=0.75) # plotting the main graph (cores * hours)
a0.fill_between(tmp_x, 0, tmp_y, color="grey",
alpha=0.45) # plotting the area below the corehours graph
for i in range(len(accum_total_time), number_of_instances * 3 +
4): # ensuring that empty months will be accounted for
accum_total_time = np.append(
accum_total_time,
accum_total_time[-1]) # filling accumulated time with most recent
if yearly_quota:
for i in range(0, int(number_of_instances)
): # not possible for the last area, hence skipping it.
monthly_used.append(accum_total_time[i * 3 + 3] -
accum_total_time[i * 3])
percentages = [0]
for i in range(len(monthly_cputime)):
if monthly_used[i] >= 1:
percentages.append(10 * (monthly_cputime[i] / monthly_used[i]))
for i in range(len(percentages)):
effarray.append(percentages[i])
a0.grid(True)
axis2 = fig.add_subplot(212)
a1legend1 = mpatches.Patch(color='Red', alpha=0.8, label="per day")
a1legend2 = mpatches.Patch(color='purple', alpha=0.8, label='per job')
a1.plot(tmp_x, delta, '.', color="purple", markersize=5,
alpha=0.35) # percentages amplified by the lower bound to
a1.legend(handles=[a1legend1, a1legend2])
plt.ylabel('Efficiency') # be more visible.
daily = []
dates = []
for i in range(int(x_start.timestamp()), int(x_end.timestamp()), 2764800):
r = D_.gather_efficiencies_for_month(
datetime.datetime.fromtimestamp(i), Data)
daily_eff_days = r[-2]
daily_eff_eff = r[-1]
r = r[:-2]
for j in range(len(r[0])):
if r[1][j] > 0:
daily.append(100 * r[1][j] / r[0][j])
dates.append(r[2][j])
formatteddates = []
for i in dates:
if len(str(i)) > 5 and "." not in str(i):
transp = str(i)[2:18]
formatteddates.append(datetime.datetime.strptime(transp, fmt))
eff_days = []
#for i in dates:
# eff_days.append(datetime.datetime.strptime(str(i)[2:18], fmt))
a1.plot(formatteddates, daily, '.', color="Red", markersize=3, alpha=0.85)
eff_distance = 0 - axis.get_ylim()[0]
a1.grid(
True
) # Creates a grid in the image to aid the viewer in visually processing the data.
a1.set_ylim([-5, 105])
if nutzergraph:
a1.set_ylim(
[0,
100]) # Usergraphs don't display anything above 100% or below 0%.
a1.set_yticks(
np.arange(0, 101, 10),
minor=True) # minor tick-lines are much thinner than regular ones
a1.set_yticks(np.arange(0, 101, 25))
a1.yaxis.set_major_formatter(mtick.PercentFormatter())
plt.xlabel('Efficiency')
plt.xlabel(' ')
a0.xaxis.tick_top()
a0.set_xlim((beginning_dt, fourteen_dt))
a1.xlim = (beginning_dt, fourteen_dt)
plt.sca(a0)
a0.yaxis.set_major_formatter(ScalarFormatter(useOffset=True))
plt.xticks(Xtics)
plt.ylabel('CPUhours')
emptylabels = []
for i in a0.get_xticklabels():
emptylabels.append(["", ""])
new_ylabels, unit = get_scaled_ylabels(a0.get_yticks())
plt.ylabel("CPUhours (" + unit + ")")
a0.set_xticklabels = emptylabels
a0.set_yticklabels(new_ylabels)
plt.sca(a1)
# dictates gap in height, left border, right border, gap in width, bottom border, top border
plt.subplots_adjust(hspace=0.03,
left=0.1,
right=0.925,
wspace=0.07,
bottom=0.035,
top=0.95)
plt.xlim(beginning_dt, fourteen_dt)
plt.xticks(Xtics)
a0.xaxis.set_major_formatter(nothing) # removes the x-tic notations
a1.xaxis.set_major_formatter(myFmt)
a1.grid(which='minor', alpha=0.2)
a1.grid(which='major', alpha=0.5)
f.set_size_inches((11, 8.5), forward=False)
## for png compression
#ram = io.BytesIO()
#plt.savefig(ram, format='png')
#ram.seek(0)
#im = Image.open(ram)
#im2 = im.convert('RGB').convert('P', palette=Image.ADAPTIVE)
#return im2
#print(tmp_x)
#print(tmp_y)
return f
#T,X,Y=read_tffile(glob( DATA_ROOT_PATH+'/data/'+ERROR_VS_TIME_PREFIX+'*' )[0])
T,Y,X=read_tffile(glob( DATA_ROOT_PATH+'/data/'+ERROR_VS_TIME_PREFIX+'*' )[0])
y_dnn=regressor.predict(input_fn=lambda :{'X': tf.convert_to_tensor(X, name='X'), 't': tf.convert_to_tensor(T, name = 't')})
logging.info('\tStart validation for '+model_name(parm))
L2_t0=dict()
L2_t1=dict()
for i,y in enumerate(y_dnn):
if i >= len(X) or i >=NUM_TIMES*NUM_SAMPLES_FOR_ERR_VS_TIME:
break
if LEARN_INVERSE:
y_pred_for_plot=np.array(y['predictions'])
tim = T[i][0]
if PLOT_ALL:
fig, (ax1,ax2,ax3)=plt.subplots(3,1,figsize=(10,6), sharex=True)
plt.sca(ax1)
plt.title('t={:5f}'.format(float(0)))
plt.plot(SAMPLING_SPACE , Y[i] , label='Real')
plt.plot(SAMPLING_SPACE , y_pred_for_plot , label='Network')
plt.ylabel('Heat profile')
plt.legend(loc='best')
plt.sca(ax2)
plt.title('t={:5f}'.format(float(tim )))
plt.plot(SAMPLING_SPACE , X[i] , label='Real' )
plt.plot(SAMPLING_SPACE , apply_heat_kernel( y_pred_for_plot, tim ) , label='Network')
plt.ylabel('Heat profile')
plt.legend(loc='best')
plt.sca(ax3)
plt.plot(SAMPLING_SPACE , y_pred_for_plot-Y[i] , label=r'$\Delta(t=0)$')
plt.plot(SAMPLING_SPACE , apply_heat_kernel( y_pred_for_plot, tim )-X[i] , label=r'$\Delta(t=$'+str(tim)+'$)$')
import numpy as np
x_min, x_max = X_train[:, 0].min() - .5, X_train[:, 0].max() + .5
y_min, y_max = X_train[:, 1].min() - .5, X_train[:, 1].max() + .5
#error in case Xs or xs
Xs = np.arange(x_min, x_max, 0.5)
fig, axes = plt.subplots(1, 3)
fig.set_size_inches(10, 6)
for i in [0, 1, 2]:
axes[i].set_aspect('equal')
axes[i].set_title('Class ' + str(i) + ' versus the rest')
axes[i].set_xlabel('Sepal length')
axes[i].set_ylabel('Sepal width')
axes[i].set_xlim(x_min, x_max)
axes[i].set_ylim(y_min, y_max)
#error here need plt.
plt.sca(axes[i])
plt.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=plt.cm.prism)
ys = (-clf.intercept_[i] - Xs * clf.coef_[i, 0]) / clf.coef_[i, 1]
plt.plot(Xs, ys, hold=True)
print(clf.predict(scaler.transform([[4.7, 3.1]])))
#[0]
print(clf.decision_function(scaler.transform([[4.7, 3.1]])))
#[[ 19.73905808 8.13288449 -28.63499119]]
from sklearn import metrics
y_train_pred = clf.predict(X_train)
print(metrics.accuracy_score(y_train, y_train_pred))
#0.821428571429
y = 4 * (X - 0.5)**2
y = y + np.random.randn(m, 1) / 10
tree_reg1 = DecisionTreeRegressor(random_state=42)
tree_reg2 = DecisionTreeRegressor(random_state=42, min_samples_leaf=10)
tree_reg1.fit(X, y)
tree_reg2.fit(X, y)
x1 = np.linspace(0, 1, 500).reshape(-1, 1)
y_pred1 = tree_reg1.predict(x1)
y_pred2 = tree_reg2.predict(x1)
fig, axes = plt.subplots(ncols=2, figsize=(10, 4), sharey=True)
plt.sca(axes[0])
plt.plot(X, y, "b.")
plt.plot(x1, y_pred1, "r.-", linewidth=2, label=r"$\hat{y}$")
plt.axis([0, 1, -0.2, 1.1])
plt.xlabel("$x_1$", fontsize=18)
plt.ylabel("$y$", fontsize=18, rotation=0)
plt.legend(loc="upper center", fontsize=18)
plt.title("No restrictions", fontsize=14)
plt.sca(axes[1])
plt.plot(X, y, "b.")
plt.plot(x1, y_pred2, "r.-", linewidth=2, label=r"$\hat{y}$")
plt.axis([0, 1, -0.2, 1.1])
plt.xlabel("$x_1$", fontsize=18)
plt.title("min_samples_leaf={}".format(tree_reg2.min_samples_leaf),
fontsize=14)
def plot_tree(t, **kwargs):
"""
Plots a tree (from module tree.py)
kwargs that do something:
nodecolors = color the nodes
leafcolors = color just the leaf nodes
title = set the title
ax = plot on the given axis
useplt = default true, if False, then ax must be specified and will
plot directly to ax.
"""
node_locs = np.zeros([t.tree_size, 2])
node_order = []
for level in xrange(1, t.tree_depth + 1):
nodes = t.dfs_level(level)
node_order.extend([x.idx for x in nodes])
node_idxs = np.array([node.idx for node in nodes])
x_intervals = np.cumsum(
np.array([0] + [node.size for node in nodes]) * 1.0 / t.size)
node_xs = x_intervals[:-1] + np.diff(x_intervals) / 2.0
node_ys = (t.tree_depth - level) * np.ones(np.shape(node_xs))
node_locs[node_idxs, :] = np.hstack(
[node_xs[:, np.newaxis], node_ys[:, np.newaxis]])
useplt = True
if "useplt" in kwargs:
if not kwargs["useplt"]:
useplt = False
ax = kwargs["ax"]
if "ax" in kwargs:
if useplt:
plt.sca(kwargs["ax"])
if "nodecolors" in kwargs:
nc = kwargs["nodecolors"]
if useplt:
plt.scatter(node_locs[:, 0],
node_locs[:, 1],
marker='.',
edgecolors='none',
c=nc,
norm=binnorm,
cmap=rbmap,
s=80)
else:
ax.scatter(node_locs[:, 0],
node_locs[:, 1],
marker='.',
edgecolors='none',
c=nc,
norm=binnorm,
cmap=rbmap,
s=80)
elif "leafcolors" in kwargs:
lc = kwargs["leafcolors"]
nonleaves = (t.tree_size - t.size)
nc = ['k'] * nonleaves
if useplt:
plt.scatter(node_locs[0:nonleaves, 0],
node_locs[0:nonleaves, 1],
edgecolors='none',
marker='.',
c=nc,
s=80)
plt.scatter(node_locs[nonleaves:, 0],
node_locs[nonleaves:, 1],
edgecolors='none',
marker='.',
c=lc,
s=80)
else:
ax.scatter(node_locs[0:nonleaves, 0],
node_locs[0:nonleaves, 1],
edgecolors='none',
marker='.',
c=nc,
s=80)
ax.scatter(node_locs[nonleaves:, 0],
node_locs[nonleaves:, 1],
edgecolors='none',
marker='.',
c=lc,
s=80)
else:
nc = 'k'
if useplt:
plt.scatter(node_locs[:, 0],
node_locs[:, 1],
marker='.',
c=nc,
s=80)
else:
ax.scatter(node_locs[:, 0],
node_locs[:, 1],
marker='.',
c=nc,
s=80)
for node in t:
if node.parent is not None:
x1, y1 = node_locs[node.idx, :]
x2, y2 = node_locs[node.parent.idx, :]
if useplt:
plt.plot((x1, x2), (y1, y2), 'r')
else:
ax.plot((x1, x2), (y1, y2), 'r')
if useplt:
plt.yticks(np.arange(0, t.tree_depth, 1))
plt.xlim([0.0, 1.0])
plt.ylim([-0.2, (t.tree_depth - 1) + 0.2])
else:
ax.set_yticks(np.arange(0, t.tree_depth, 1))
ax.set_xlim([0.0, 1.0])
ax.set_ylim([-0.2, (t.tree_depth - 1) + 0.2])
if "title" in kwargs:
if useplt:
plt.title(kwargs["title"])
else:
ax.set_title(kwargs["title"])
if "nodelocs" in kwargs:
return node_locs
gape_array = hf5.root.ancillary_analysis.gapes[:]
ltp_array = hf5.root.ancillary_analysis.ltps[:]
gape_array_long = gape_array.reshape(\
(gape_array.shape[0],np.prod(gape_array.shape[1:3]),gape_array.shape[-1]))
#gape_array_long = gape_array_long[:,:,2000:4500]
gape_array_really_long = gape_array.reshape(\
(np.prod(gape_array.shape[:3]),gape_array.shape[3]))
ltp_array_long = ltp_array.reshape(\
(ltp_array.shape[0],np.prod(ltp_array.shape[1:3]),ltp_array.shape[-1]))
ltp_array_long = ltp_array_long[:, :, 2000:4500]
fig, ax = plt.subplots(2, 2)
plt.sca(ax[0, 0])
dat_imshow(gape_array_long[0, :, :])
plt.sca(ax[0, 1])
dat_imshow(gape_array_long[1, :, :])
plt.sca(ax[1, 0])
dat_imshow(ltp_array_long[0, :, :])
plt.sca(ax[1, 1])
dat_imshow(ltp_array_long[1, :, :])
plt.show()
# Find bout length
this_array = gape_array_long[0, :, :]
def calc_bout_duration(gape_array):
gape_array[gape_array < 0.5] = 0
if ts[0] < 2:
aux = ax.imshow(ts[1], origin='lower', zorder=5, cmap=current_cmap)
else:
aux = ax.imshow(ts[1], origin='lower', zorder=5, cmap=current_cmap2)
ax.plot(_bo_xs[:, 0], _bo_xs[:, 1], '^y', zorder=10)
ax.plot(i_pos[0], i_pos[1], '^y', zorder=10)
CS = ax.contour(ts[1], colors='k', alpha=0.6, linewidths=1.3, zorder=10)
divider = make_axes_locatable(ax)
cax = divider.append_axes('right', size='5%', pad=0.05)
fig.colorbar(aux, cax=cax, orientation='vertical')
ax.grid(True, zorder=0, color="white")
ax.clabel(CS, inline=1, fontsize=12)
ax.set_facecolor('#eaeaf2')
ax.set_xlabel("x (m)", fontsize=17)
ax.set_xticks(xticks)
ax.set_xticklabels(xnticks, fontsize=17)
ax.set_yticks(yticks)
ax.set_yticklabels(ynticks, fontsize=0)
if ts[0] == 0:
ax.set_title("$\\alpha_1(x)$")
elif ts[0] == 1:
ax.set_title("$\\alpha_2(x)$")
elif ts[0] == 2:
ax.set_title("$\\sigma_1(x)$")
else:
ax.set_title("$\\sigma_2(x)$")
plt.sca(axs[0])
plt.yticks(yticks, labels=ynticks, fontsize=17)
plt.ylabel("y (m)", fontsize=17)
plt.show(block=True)
def elements(elem,*args,**kwargs):
"""
NAME:
elements
PURPOSE:
make a plot of measurements of the elemental abundances vs. atomic number
INPUT:
elem - dictionary with elemental abundances relative to H
wrtFe= (True) if True, plot elements wrt Fe on the left Y
inclwrtH= (True) if True, indicate what X/H is on the right Y
bovy_plot.bovy_plot args and kwargs
OUTPUT:
plot to output
HISTORY:
2015-03-10 - Written - Bovy (IAS)
"""
# Process the input dictionary
xs= []
names= []
ys= []
wrtFe= kwargs.pop('wrtFe',True)
for el in elem:
try:
xs.append(atomic_number(el))
except KeyError: # ignore things that aren't known elements
continue
names.append(r'$\mathrm{%s}$' % el.lower().capitalize())
try:
if not wrtFe: raise KeyError
ys.append(elem[el]-elem['Fe'])
except KeyError:
ys.append(elem[el])
wrtFe= False
xs= numpy.array(xs,dtype='int')
ys= numpy.array(ys)
names= numpy.array(names)
# sort
sindx= numpy.argsort(xs)
xs= xs[sindx]
ys= ys[sindx]
names= names[sindx]
# add second y axis?
inclwrtH= kwargs.pop('inclwrtH',True)
if wrtFe:
feh= elem['Fe']
ylabel= kwargs.pop('ylabel',r'$[\mathrm{X/Fe}]$')
else:
ylabel= kwargs.pop('ylabel',r'$[\mathrm{X/H}]$')
if not kwargs.get('overplot',False):
bovy_plot.bovy_print(fig_width=7.,fig_height=4.)
basezorder= kwargs.pop('zorder',0)
yrange=kwargs.pop('yrange',[-0.5,0.5])
ls= kwargs.pop('ls','-')
lw= kwargs.pop('lw',0.25)
bovy_plot.bovy_plot(xs,ys,*args,
ylabel=ylabel,
xrange=[4,numpy.amax(xs)+2],
yrange=yrange,zorder=2+basezorder,ls=ls,lw=lw,
**kwargs)
pyplot.xticks(list(xs),names)
pyplot.tick_params(axis='x',labelsize=11.)
if wrtFe and inclwrtH:
bovy_plot.bovy_plot([4,numpy.amax(xs)+2],[0.,0.],'-',lw=2.,
color='0.65',overplot=True,zorder=basezorder)
ax= pyplot.gca()
ax2= ax.twinx()
ax2.set_ylim(yrange[0]+feh,yrange[1]+feh)
ax2.set_ylabel(r'$[\mathrm{X/H}]$')
pyplot.sca(ax2)
bovy_plot._add_ticks(yticks=True,xticks=False)
return None
def __init__(self, data):
""" class initializer """
# store relevant dictionary entries into convenient local variables
self.ultra_interp = data['ultra_interp']
self.ofdisp = data['ofdisp']
self.ult_no_frames = data['ult_no_frames']
self.ult_time = data['ult_time']
self.ult_period = self.ult_time[1] - self.ult_time[0]
self.wav_time = data['wav_time']
self.wav_data = data['wav_data'] / max(abs(data['wav_data']))
self.wav_fs = data['wav_fs']
self.wav_num_samples = len(self.wav_data)
# Visualization options
self.flow_dir = FlowDir.Horizontal
self.flow_polarity = 1
self.flow_half_field = 0
self.flow_dir_label = 'Horizontal'
self.flow_pol_label = ''
self.skip_increment = 3
self.pos_ylim_del = 1
self.vel_ylim_del = 10
self.pos_ylim_min = 1
self.pos_ylim_max = 200
self.vel_ylim_min = 10
self.vel_ylim_max = 2000
# TODO implement vector scaling as a user parameter
self.scaling = 1.0
# visualize registration as quiver plot
self.quiver_x_range = range(1, self.ultra_interp[0].shape[0])
self.quiver_y_range = range(1, self.ultra_interp[0].shape[1])
self.xx, self.yy = np.meshgrid(self.quiver_x_range,
self.quiver_y_range)
self.x_indices_full, self.y_indices_full = np.meshgrid(
np.arange(0, self.xx.shape[0]), np.arange(0, self.xx.shape[1]))
self.x_indices_half, self.y_indices_half = np.meshgrid(
np.arange(0, int(self.xx.shape[0] / 2)),
np.arange(0, self.xx.shape[1]))
self.x_indices_plot, self.y_indices_plot = np.meshgrid(
np.arange(0, self.xx.shape[0], self.skip_increment),
np.arange(0, self.xx.shape[1], self.skip_increment))
self.quiver_colors = np.empty(shape=np.shape(self.x_indices_plot))
self.quiver_colors.fill(1)
# TODO: A lazy hack to force the quiver colormapping to work (it autoscales to the values)
self.quiver_colors[0, 0] = 0
self.quiver_plot_halfsize = math.ceil(
self.x_indices_plot.shape[0] / 2) * self.x_indices_plot.shape[1]
self.x_indices = self.x_indices_full
self.y_indices = self.y_indices_full
#TODO set the default figure size to be some sensible proportion of the screen real estate
self.fig = plt.figure(figsize=(10, 8))
self.ax_quiver = self.fig.add_axes([0.1, 0.6, 0.4, 0.4])
plt.sca(self.ax_quiver)
#plt.set_cmap('Greys')
self.im = self.ax_quiver.imshow(self.ultra_interp[0])
self.quiver = plt.quiver(self.yy[self.x_indices_plot,
self.y_indices_plot],
self.xx[self.x_indices_plot,
self.y_indices_plot],
self.ofdisp[0]['of'][0][self.y_indices_plot,
self.x_indices_plot],
self.ofdisp[0]['of'][1][self.y_indices_plot,
self.x_indices_plot],
self.quiver_colors,
cmap=cm.bwr,
scale_units='xy',
scale=0.5,
angles='xy')
# compute the velocity and position using the trimmed mean approach
self.vel = np.empty((self.ult_no_frames - 1, 2))
self.pos = np.empty(
(self.ult_no_frames - 2, 2)
) # integration of velocity leaves us with N - 1, we pad afterwards
# Add a compass plot to visualize the consensus vector
self.ax_compass = self.fig.add_axes([0.7, 0.7, 0.2, 0.2],
projection='polar')
self.ax_compass.set_ylim(0, 100)
# create velocity plot
self.ax_vel = self.fig.add_axes([0.1, 0.5, 0.8, 0.1])
self.line_vel, = plt.plot(self.ult_time[0:self.ult_no_frames - 1],
np.zeros(self.ult_no_frames - 1) *
np.nan) #self.vel[:, self.flow_dir.value])
plt.axhline(linewidth=1, color='k')
# TODO: Setting the xlim changes the tick scaling but not the data scale
self.ax_vel.set_xlim([0.0, self.ult_time[-1]])
self.ax_vel.set_ylim([-1e2, 1e2])
self.title_vel = self.ax_vel.set_title("Velocity (" +
self.flow_dir_label +
" in Video)")
self.ax_vel.set_ylabel("velocity (mm/s)")
self.ax_vel.tick_params(bottom=False)
# create position plot
self.ax_pos = self.fig.add_axes([0.1, 0.3, 0.8, 0.1])
self.line_pos, = plt.plot(self.ult_time[0:self.ult_no_frames - 1],
np.zeros(self.ult_no_frames - 1) *
np.nan) #self.pos[:, self.flow_dir.value])
plt.axhline(linewidth=1, color='k')
# TODO: Setting the xlim changes the tick scaling but not the data scale
self.ax_pos.set_xlim([0.0, self.ult_time[-1]])
self.ax_pos.set_ylim([-2e1, 2e1])
self.title_pos = self.ax_pos.set_title("Position (" +
self.flow_dir_label +
" in Video)")
self.ax_pos.set_ylabel("relative position (mm)")
self.ax_pos.set_xlabel("time (s)")
# cache the axis for faster rendering
# create audio plot
self.ax_audio = self.fig.add_axes([0.1, 0.1, 0.8, 0.1])
self.line_audio, = plt.plot(self.wav_time, self.wav_data)
self.ax_audio.set_xlim([0.0, self.ult_time[-1]])
self.ax_audio.set_ylim([-1, 1])
plt.axhline(linewidth=1, color='k')
self.ax_audio.set_title("Audio")
self.frame_index = 0
# connect the callbacks
self.cid_scroll = self.fig.canvas.mpl_connect('scroll_event',
self.mouse_scroll)
self.fig.canvas.mpl_connect('key_press_event', self.key_press)
# TODO: Need to stop execution upon closing the matplotlib figure (calling exit(0) via an on_close callback predictably generates an error)
#self.fig.canvas.mpl_connect('close_event', self.on_close)
# cache the axis for faster rendering
self.fig.canvas.draw()
self.ax_quiver_bg = self.fig.canvas.copy_from_bbox(self.ax_quiver.bbox)
self.ax_vel_clear_bg = self.fig.canvas.copy_from_bbox(self.ax_vel.bbox)
self.ax_pos_clear_bg = self.fig.canvas.copy_from_bbox(self.ax_pos.bbox)
self.ax_audio_bg = self.fig.canvas.copy_from_bbox(self.ax_audio.bbox)
self.ax_compass_bg = self.fig.canvas.copy_from_bbox(
self.ax_compass.bbox)
# compute the kinematics and update the filled plots
self.pos, self.vel = compute_kinematics(self.ofdisp, self.ult_time,
self.ult_no_frames,
self.x_indices, self.y_indices,
self.ult_period, self.scaling)
self.line_vel.set_ydata(self.flow_polarity *
self.vel[:, self.flow_dir.value])
self.line_pos.set_ydata(self.flow_polarity *
self.pos[:, self.flow_dir.value])
self.fig.canvas.draw()
self.ax_vel_filled_bg = self.fig.canvas.copy_from_bbox(
self.ax_vel.bbox)
self.ax_pos_filled_bg = self.fig.canvas.copy_from_bbox(
self.ax_pos.bbox)
# now that we have cached the background add the varying plot features (image, quiver, time markers)
plt.sca(self.ax_quiver)
self.quiver_frame_label = plt.text(1,
self.quiver_y_range[-1] * 0.15,
"1",
fontsize=48,
color="white")
plt.sca(self.ax_vel)
self.point_vel, = plt.plot(self.ult_time[0],
self.vel[0, self.flow_dir.value],
marker="o",
ls="",
color="r")
plt.sca(self.ax_pos)
self.point_pos, = plt.plot(self.ult_time[0],
self.pos[0, self.flow_dir.value],
marker="o",
ls="",
color="r")
plt.sca(self.ax_audio)
self.marker_line_audio, = plt.plot(
[self.wav_time[0], self.wav_time[0]], [-1, 1], color="r")
plt.sca(self.ax_compass)
self.marker_line_compass, = plt.plot(
[0, 1], [0, 1], color="r"
) #("", xy=(angle, radius), xytext=(0, 0), arrowprops=dict(arrowstyle="<-", color='k'))
self.fig.canvas.draw()
plt.show()
plt.scatter(X[:, 0], X[:, 1], c="k")
plt.title("$Initial$", fontsize=18)
plt.xlabel("X1", fontsize=15)
plt.ylabel("X2", fontsize=15, rotation=0)
plt.savefig("./K-means_Initial.png")
plt.show()
distances = distanceToCentroids(X, centroids)
centroids, centroid_points = getNewCentroid(X, distances, num_centroids)
fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(14, 9))
for i in range(num_iterations):
plt.sca(axes[i // 2, i % 2])
for idx, centroid in enumerate(centroids, 1):
x1, x2 = [], []
for a, b in centroid_points[idx - 1]:
x1.append(a)
x2.append(b)
plt.plot(centroid[0],
centroid[1],
marker="^",
c=colors[idx - 1],
markersize=12,
label=f'k{idx}')
plt.scatter(x1, x2, s=15, c=colors[idx - 1], label=f'Centroid: {idx}')
def update(self, event):
if event.inaxes != self.axes4:
return
if event.xdata != None:
x = (int(event.xdata)+self.imageWidth//2)%self.imageWidth
y = (int(event.ydata)+self.imageHeight//2)%self.imageHeight
plt.sca(self.axes5)
plt.cla()
waveImg = numpy.zeros((self.imageHeight,self.imageWidth))
waveImg[y,x] = 1
plt.imshow(numpy.real(fftpack.ifft2(waveImg)), cmap='gray')
if not self.bCtrlPressed:
bNeedUpdate = False
if self.samples[y,x] != self.fftImage[y,x] and self.mouseButton == 1: #left button
bNeedUpdate = True
self.samples[y,x] = self.fftImage[y,x]
self.samplePoints[(y-self.imageHeight//2)%self.imageHeight,(x-self.imageWidth//2)%self.imageWidth,0] = 1
self.samplePoints[(y-self.imageHeight//2)%self.imageHeight,(x-self.imageWidth//2)%self.imageWidth,3] = 1
elif self.samples[y,x] != numpy.complex(0.0,0.0) and self.mouseButton == 3: #right button
bNeedUpdate = True
self.samples[y,x] = numpy.complex(0.0,0.0)
self.samplePoints[(y-self.imageHeight//2)%self.imageHeight,(x-self.imageWidth//2)%self.imageWidth,0] = 0
self.samplePoints[(y-self.imageHeight//2)%self.imageHeight,(x-self.imageWidth//2)%self.imageWidth,3] = 0
if bNeedUpdate:
plt.sca(self.axes4)
plt.cla()
p = plt.imshow(self.fftImageForPlot, cmap='gray')
p.set_clim(self.fftMean-self.fftStd,self.fftMean+self.fftStd)
plt.imshow(self.samplePoints)
plt.sca(self.axes3)
plt.cla()
plt.imshow(numpy.real(fftpack.ifft2(self.samples)), cmap='gray')
else:
for xi in range(x-self.imageWidth//32, x+self.imageWidth//32):
for yi in range(y-self.imageWidth//32, y+self.imageWidth//32):
if xi>=self.imageWidth:
xx = xi-self.imageWidth
else:
xx = xi
if yi>=self.imageHeight:
yy = yi-self.imageHeight
else:
yy = yi
if self.mouseButton == 1: #left button
self.samples[yy,xx] = self.fftImage[yy,xx]
self.samplePoints[(yy-self.imageHeight//2)%self.imageHeight,(xx-self.imageWidth//2)%self.imageWidth,0] = 1
self.samplePoints[(yy-self.imageHeight//2)%self.imageHeight,(xx-self.imageWidth//2)%self.imageWidth,3] = 0.7
elif self.mouseButton == 3: #right button
self.samples[yy,xx] = numpy.complex(0.0,0.0)
self.samplePoints[(yy-self.imageHeight//2)%self.imageHeight,(xx-self.imageWidth//2)%self.imageWidth,0] = 0
self.samplePoints[(yy-self.imageHeight//2)%self.imageHeight,(xx-self.imageWidth//2)%self.imageWidth,3] = 0
plt.sca(self.axes4)
plt.cla()
plt.imshow(self.samplePoints)
plt.sca(self.axes3)
plt.cla()
plt.imshow(numpy.real(fftpack.ifft2(self.samples)), cmap='gray')
self.fig.canvas.draw()
def ftAllSkyImage(allSkyImage, options, labelOptions, stationLocation, lVec,
mVec):
"""Summary
Args:
allSkyImage (TYPE): Description
options (TYPE): Description
labelOptions (TYPE): Description
stationLocation (TYPE): Description
lVec (TYPE): Description
mVec (TYPE): Description
Returns:
TYPE: Description
"""
plotOptions = options['plottingOptions']
logPlot, skyObjColor, gridThickness, backgroundColor, foregroundColor, radialLabelAngle, colorBar, obsSite, outputFolder, maxPercentile, minPercentile, figureShape, fontSizeFactor = plotOptions[
'logPlot'], plotOptions['skyObjColor'], plotOptions[
'gridThickness'], plotOptions['backgroundColor'], plotOptions[
'foregroundColor'], plotOptions[
'radialLabelAngle'], plotOptions['colorBar'], options[
'stationID'], plotOptions['outputFolder'], plotOptions[
'maxPercentile'], plotOptions[
'minPercentile'], plotOptions[
'figureShape'], plotOptions[
'fontSizeFactor']
dateTime, rcuMode, subband, frequency, polarity, figNum = labelOptions
if fontSizeFactor is None:
fontSizeFactor = figureShape[0] / 18.
if options['rfiMode']:
fontSizeFactor = min(fontSizeFactor, 0.66)
telescopeLoc = astropy.coordinates.EarthLocation(
lon=stationLocation[0] * u.deg,
lat=stationLocation[1] * u.deg,
height=stationLocation[2] * u.m)
if len(dateTime) == 15:
dateTime = str(datetime.datetime.strptime(dateTime, '%Y%m%d-%H%M%S'))
obsTime = astropy.time.Time(dateTime)
if plotOptions['plotSkyObjects']:
knownSources, referenceObject = getSkyObjects(options, obsTime,
telescopeLoc)
altAzRef = astropy.coordinates.AltAz(obstime=obsTime,
location=telescopeLoc)
altAzObjects = [
skyObj.transform_to(altAzRef) for skyObj in referenceObject
]
gs = mpl.gridspec.GridSpec(1, 2, width_ratios=[18, 1])
gs2 = mpl.gridspec.GridSpec(1, 2, width_ratios=[18, 1])
fig = plt.figure(figNum, figsize=(figureShape[0], figureShape[1]))
fig.patch.set_facecolor(backgroundColor)
plt.suptitle(
'LOFAR mode {0}{1} all sky plot at {2}MHz (sb{3}) for {4}\n'.format(
rcuMode, polarity, round(frequency / 1e6, 2), subband, obsSite),
fontsize=int(28 * fontSizeFactor),
color=foregroundColor) #, va = 'top')
plt.rcParams["text.color"] = foregroundColor
plt.rcParams["axes.labelcolor"] = foregroundColor
plt.rcParams["xtick.color"] = foregroundColor
plt.rcParams["ytick.color"] = foregroundColor
plt.rcParams['axes.edgecolor'] = foregroundColor
plt.rcParams['axes.linewidth'] = gridThickness
axImage = fig.add_subplot(gs[0], label='ax_image')
axImage.axis('off')
global vmaxCache
global vminCache
if logPlot:
allSkyImageLog = np.log10(allSkyImage)
if isinstance(vmaxCache, float):
vmaxVar = np.log10(vmaxCache)
vminVar = np.log10(vminCache)
else:
vmaxVar = np.nanpercentile(allSkyImageLog, maxPercentile)
vminVar = np.nanpercentile(allSkyImageLog, minPercentile)
vminCache.append(vminVar)
vmaxCache.append(vmaxVar)
vminVar = np.mean(vminCache)
vmaxVar = np.mean(vmaxCache)
pltIm = axImage.imshow(allSkyImageLog,
cmap='jet',
label='ax_image',
interpolation='lanczos',
vmax=vmaxVar,
vmin=vminVar)
else:
if isinstance(vmaxCache, float):
vmaxVar = vmaxCache
vminVar = vminCache
else:
vmaxVar = np.nanpercentile(allSkyImage, maxPercentile)
vminVar = np.nanpercentile(allSkyImage, minPercentile)
vminCache.append(vminVar)
vmaxCache.append(vmaxVar)
vminVar = np.mean(vminCache)
vmaxVar = np.mean(vmaxCache)
pltIm = axImage.imshow(allSkyImage,
cmap='jet',
label='ax_image',
interpolation='lanczos',
vmax=vmaxVar,
vmin=vminVar)
axImage.axis('off')
if colorBar:
axColorBar = plt.subplot(gs[1])
colorBarObj = plt.colorbar(pltIm, axColorBar)
axColorBar.tick_params(which='minor', length=2)
axColorBar.tick_params(which='major', length=4, width=1)
axColorBar.yaxis.set_minor_locator(mpl.ticker.AutoMinorLocator(10))
if options['rfiMode']:
cbCursor = axColorBar.plot([0, 1], [0, 0], 'k-')
pltObj = fig.add_subplot(gs2[0], label='ax', polar=True)
axImage.set_xlim((0, lVec.size))
axImage.set_ylim((lVec.size, 0))
pltObj.set_theta_zero_location("N")
pltObj.set_theta_direction(1)
if plotOptions['plotSkyObjects']:
plotStatus = [[
True,
__plotSkyObject(axImage,
skyObj,
lVec.size,
skyObjColor,
knownSources[idx],
fontSizeFactor=fontSizeFactor)
] if (skyObj.alt.deg > 20.) or
(options['rfiMode'] and skyObj.alt.deg > -10.) else [
False,
__plotSkyObject(axImage,
skyObj,
lVec.size,
skyObjColor,
knownSources[idx],
offset=True,
fontSizeFactor=fontSizeFactor)
] for idx, skyObj in enumerate(altAzObjects)]
legend = axImage.legend(loc=8,
bbox_to_anchor=(0.5, -0.128),
ncol=4,
framealpha=0.0,
fontsize=int(14 * fontSizeFactor),
title=str(obsTime)[:-4])
legend.get_title().set_fontsize(str(int(22 * fontSizeFactor)))
for idx, skyText in enumerate(legend.get_texts()):
if not plotStatus[idx][0]:
plt.setp(skyText, color='red')
elif options['rfiMode'] and altAzObjects[
idx].alt.deg > -10. and altAzObjects[idx].alt.deg < 20.:
plt.setp(skyText, color='orange')
radii = []
if plotOptions['graticule']:
for radius in range(0, 90, 15): # r grid at 15 degree intervals
radii.append(
180 * np.cos(radius * np.pi / 180)
) # plot the radii so as to display as an orthographic grid
pltObj.set_rgrids(radii)
if radialLabelAngle: # you would not want to put y ticks on 0 anyhow as it would be messy
yLabel = [
'', '15' + u'\xb0', '30' + u'\xb0', '45' + u'\xb0',
'60' + u'\xb0', '75' + u'\xb0'
]
pltObj.set_yticklabels(yLabel, color=foregroundColor)
print(radialLabelAngle)
pltObj.set_rlabel_position(radialLabelAngle)
else:
yLabel = []
pltObj.set_yticklabels(yLabel)
thetaticks = np.arange(0, 360, 45)
pltObj.set_thetagrids(thetaticks,
weight='bold',
color=foregroundColor,
fontsize=int(18 * fontSizeFactor))
pltObj.tick_params('x', pad=0, rotation='auto')
pltObj.grid(False,
'both',
color=backgroundColor,
linewidth=gridThickness)
pltObj.patch.set(alpha=0.0)
plt.sca(axImage)
if options['rfiMode']:
from rfiPlotter import onclick, hover, onaxesleave
annot = pltObj.annotate("",
xy=(0, 0),
xytext=(15, 15),
textcoords="offset points",
bbox=dict(boxstyle="round", fc="w"),
arrowprops=dict(arrowstyle="->"))
annot.set_visible(False)
global informationArr
informationArr = {
'pltObj': pltObj,
'annot': annot,
'pixels': options['imagingOptions']['pixelCount'][0],
'cbCursor': cbCursor,
'rawdata': allSkyImage,
'axColorBar': axColorBar
}
onclickLambda = lambda clickEvent: onclick(clickEvent, **informationArr
)
onaxesleaveLambda = lambda clickEvent: onaxesleave(
clickEvent, **informationArr)
hoverLambda = lambda event: hover(event, **informationArr)
fig.canvas.mpl_connect("motion_notify_event", hoverLambda)
fig.canvas.mpl_connect('button_press_event', onclickLambda)
fig.canvas.mpl_connect('axes_leave_event', onaxesleaveLambda)
plt.ioff()
plt.show()
plt.pause(1)
elif plotOptions['displayImages']:
plt.ioff()
plt.show()
plt.pause(1)
outputFolder += '/sb_{0}/'.format(subband)
if not os.path.exists(outputFolder):
os.makedirs(outputFolder)
plotFilename = "{7}{0}_{1}_sb{2}_mode{3}{4}_{5}_{6}MHz.png".format(
dateTime, obsSite, subband, rcuMode, polarity,
options['imagingOptions']['method'].replace('/', '-'),
int(frequency / 1e6), outputFolder)
plotFilename = plotFilename.replace(' ', '_').replace(':', '')
print("Saving output to {0}".format(plotFilename))
fig.savefig(plotFilename, facecolor=fig.get_facecolor(), edgecolor='none')
plt.close(figNum)
return plotFilename
title('Lesion')
xlabel('Time (seconds)')
ylabel('Progression (cm)')
xlim(0, 3)
plt.subplot(1, 2, 2)
[plot_trial_parameter(a, u1t, color='b') for a in cg]
title('Sham/Control')
xlabel('Time (seconds)')
xlim(0, 3)
### PROGRESSION BEFORE AND AFTER COMPARISON ###
fig, axs = plt.subplots(6, 2, True, True)
axs = axs.ravel()
subjects = [0, 1, 2, 3, 4, 5, 8, 7, 10, 11, 12, 13]
for i in range(len(subjects)):
plt.sca(axs[i])
plot_progression(subjects[i], label=False)
if i == 0:
title('Lesion')
legend(bbox_to_anchor=(0, 0, 0.5, 1), bbox_transform=gcf().transFigure)
if i == 1:
title('Sham/Control')
if i == 4:
ylabel('Progression (cm)')
if i > 9:
xlabel('Time (seconds)')
[plot(trial[:, 0], 'r') for trial in u1t[lg]]
[plot(trial[:, 0], 'b') for trial in u1t[cg]]
### RAND ###
def draw(data):
"""
画图
:return:
"""
types = data['types']
cpu_time = data['cpu_time']
cpu = data['cpu']
mem = data['mem']
jvm = data['jvm']
io_time = data['io_time']
io = data['io']
disk = data['disk']
rec = data['rec']
trans = data['trans']
net = data['nic']
length = len(cpu_time)
io_length = len(io_time)
net_length = len(net)
if min(length, io_length, net_length) < 7: # 画图的最小刻度为7，故必须大于7个数据
logger.error('数据太少，请稍后再试')
raise Exception('当前数据太少，请稍后再试')
index = [[], []] # x轴坐标，一个是CPU和内存的x轴，一个是IO的x轴
labels = [[], []] # x轴刻度
delta = length / 6 # x轴每个坐标间隔
io_delta = io_length / 6 # x轴每个坐标间隔，IO
for i in range(6):
index[0].append(int(i * delta))
labels[0].append(utc2local(cpu_time[int(i * delta)]))
if types == 'system':
index[1].append(int(i * io_delta))
labels[1].append(utc2local(io_time[int(i * io_delta)]))
# 添加最后一个时刻的数据
index[0].append(length - 1)
labels[0].append(utc2local(cpu_time[length - 1]))
# 数据的开始时间和结束时间，用于计算图中展示数据的总时长
cpu_start_time = time.mktime(
datetime.datetime.strptime(cpu_time[0].split('.')[0],
'%Y-%m-%dT%H:%M:%S').timetuple())
cpu_end_time = time.mktime(
datetime.datetime.strptime(cpu_time[-1].split('.')[0],
'%Y-%m-%dT%H:%M:%S').timetuple())
if types == 'system':
# 添加最后一个时刻的数据
index[1].append(io_length - 1)
labels[1].append(utc2local(io_time[io_length - 1]))
# 用于计算总时长
io_start_time = time.mktime(
datetime.datetime.strptime(io_time[0].split('.')[0],
'%Y-%m-%dT%H:%M:%S').timetuple())
io_end_time = time.mktime(
datetime.datetime.strptime(io_time[-1].split('.')[0],
'%Y-%m-%dT%H:%M:%S').timetuple())
fig = plt.figure('figure', figsize=(20, 20))
ax1 = plt.subplot(4, 1, 1)
ax2 = plt.subplot(4, 1, 2)
ax3 = plt.subplot(4, 1, 3)
ax4 = plt.subplot(4, 1, 4)
# 画CPU使用率
plt.sca(ax1)
plt.plot(cpu, color='r', linewidth=0.3, label='CPU')
plt.grid()
plt.xlim(0, len(cpu))
plt.ylim(0, 100)
plt.title(
'CPU(%), max:{:.2f}%, average:{:.2f}%, duration:{:.1f}h'.format(
max(cpu),
sum(cpu) / len(cpu), (cpu_end_time - cpu_start_time) / 3600),
size=12)
plt.xticks(index[0], labels[0])
plt.margins(0, 0)
# 画内存使用大小
plt.sca(ax2)
plt.plot(mem, color='r', linewidth=1, label='Memory')
plt.title('Free Memory(G), min:{:.2f}G, duration:{:.1f}h'.format(
min(mem), (cpu_end_time - cpu_start_time) / 3600),
size=12)
plt.grid()
plt.xlim(0, len(mem))
plt.ylim(0, max(mem) + 1)
plt.xticks(index[0], labels[0])
plt.margins(0, 0)
# 画磁盘IO
plt.sca(ax3)
plt.plot(io, color='r', linewidth=1, label='%util')
plt.grid()
plt.xlim(0, len(io))
plt.ylim(0, max(io))
plt.title('IO({}), max:{:.2f}%, duration:{:.1f}h'.format(
disk, max(io), (io_end_time - io_start_time) / 3600),
size=12)
plt.xticks(index[1], labels[1])
plt.margins(0, 0)
# 画带宽图
plt.sca(ax4)
plt.plot(rec, color='orange', linewidth=0.5, label='rMbs')
plt.plot(trans, color='blue', linewidth=0.5, label='tMbs')
plt.legend(loc='upper left')
plt.grid()
plt.xlim(0, len(rec))
plt.ylim(0, max(max(rec), max(trans)))
plt.title(
'NetWork, Rmax:{:.2f}Mb/s, Tmax:{:.2f}Mb/s, NetWork:{:.2f}%, duration:{:.1f}h'
.format(max(rec), max(trans), max(net),
(cpu_end_time - cpu_start_time) / 3600),
size=12)
plt.xticks(index[0], labels[0])
plt.margins(0, 0)
ax_net = ax4.twinx()
plt.sca(ax_net)
plt.plot(net, color='red', linewidth=0.5, label='%net')
plt.legend(loc='upper right')
plt.ylim(0, max(net))
else:
fig = plt.figure('figure', figsize=(20, 10))
ax1 = plt.subplot(2, 1, 1)
ax2 = plt.subplot(2, 1, 2)
# 画CPU使用率
plt.sca(ax1)
plt.plot(cpu, color='r', linewidth=0.3, label='CPU')
plt.grid()
plt.xlim(0, len(cpu))
plt.ylim(0, 100)
plt.title(
'CPU(%), max:{:.2f}%, average:{:.2f}%, duration:{:.1f}h'.format(
max(cpu),
sum(cpu) / len(cpu), (cpu_end_time - cpu_start_time) / 3600),
size=12)
plt.xticks(index[0], labels[0])
plt.margins(0, 0)
# 画内存使用大小和jvm
plt.sca(ax2)
plt.plot(mem, color='r', linewidth=1, label='Memory')
if sum(jvm) == 0: # 如果没有jvm数据，则不画jvm
plt.title('Memory(G), max:{:.2f}G, duration:{:.1f}h'.format(
max(mem), (cpu_end_time - cpu_start_time) / 3600),
size=12)
else:
plt.plot(jvm, color='b', linewidth=1, label='JVM')
plt.legend(loc='upper right')
plt.title(
'Memory(G) max:{:.2f}G, JVM(G) max:{:.2f}G, duration:{:.1f}h'.
format(max(mem), max(jvm),
(cpu_end_time - cpu_start_time) / 3600),
size=12)
plt.grid()
plt.xlim(0, len(mem))
plt.ylim(0, max(max(mem), max(jvm)) + 1)
plt.xticks(index[0], labels[0])
plt.margins(0, 0)
image_byte = BytesIO()
fig.savefig(image_byte, format='png', bbox_inches='tight') # 把图像保存成二进制
img = base64.encodebytes(image_byte.getvalue()).decode() # 二进制图片流转base64编码
plt.close()
return {'img': img}
criterion='mae',
n_jobs=-1,
bootstrap=True)
#warm_start=False,
#max_leaf_nodes = 30)
rfr.fit(X_train, y_train)
predictions = rfr.predict(X_test)
#plt.scatter(y_test,predictions)
print('MAE:', metrics.mean_absolute_error(y_test, predictions))
print('MSE:', metrics.mean_squared_error(y_test, predictions))
print('RMSE:', np.sqrt(metrics.mean_squared_error(y_test, predictions)))
print('MAPE:', mean_absolute_percentage_error(y_test, predictions))
#columns = net_profit_percent.columns
#print (sorted(zip(map(lambda x: round(x, 4), rfr.feature_importances_), columns),reverse=True))
#subplots method of matplotlib
fig, axes = plt.subplots(nrows=2, ncols=1)
axes[0].scatter(y_test, predictions)
plt.sca(axes[1]) #Use the pyplot interface to change just one subplot
plt.xticks(range(X_train.shape[1]), X_train.columns, color='r')
axes[1].bar(range(X_train.shape[1]),
rfr.feature_importances_,
color='b',
align='center')
#implement the profit_prediction algorihtm on pop skus
second.drop('sku_id', axis=1, inplace=True)
pop_predictions = rfr.predict(second)
def plot_params(self,
pres_min,
pres_max,
param1,
param2='z',
delta=True,
day_range=None,
saveas=None,
plottype='average'):
"""
Plot a given parameter's average value over a given depth range
@params
param1 - An ArgoProfile attribute, on the gen_attrlist
param2 - An ArgoProfile attribute, on the gen_attrlist
pres_min - Minimum pressure to plot from
pres_max - Maximum pressure to plot from
delta - Whether to plot absolute value or deviation from average
day_range - Day range to plot (last day in list is top value for location/parameter plot)
saveas - String file path to save this plot to, as a png file, if any (otherwise call plt.show())
"""
# Plan not to display if saving, by turning interactive mode off (or on if planning to show)
if saveas != None: plt.ioff()
else: plt.ion()
# Generate day range if needed. Format to sorted just in case.
if day_range is None:
day_range = self.any_match_days()
day_range = sorted(day_range)
# Generate pressure range
pres_range = np.arange(pres_min, pres_max, 1.0)
# PREPARE AXES
f = plt.figure(figsize=(13, 8))
gs = gridspec.GridSpec(6, self.n_floats, figure=f)
map_axes = [plt.subplot(gs[0:2, i]) for i in range(self.n_floats)]
par_axes = [plt.subplot(gs[2:4, i]) for i in range(self.n_floats)]
day_axis = plt.subplot(gs[4:6, :])
# Add axis labels and format shape to place nicer when saving without adjustments.
for par_axis in par_axes:
par_axis.set_xlabel(ap.attr_to_name(param1))
par_axis.set_ylabel(ap.attr_to_name(param2))
day_axis.set_xlabel("Date")
day_axis.set_ylabel(str(param1) + " " + plottype)
plt.tight_layout()
plt.subplots_adjust(top=0.94,
bottom=0.203,
left=0.081,
right=0.98,
hspace=0.751,
wspace=0.481)
# AXES : map_axes : Float location on map with SLA
# For each argo float
for argofloat, map_axis, wmo_id in zip(self.floats, map_axes,
self.wmo_ids):
# a. Draw the map
map = argofloat.aviso_map_plot(day_range[-1], map_axis)
# b. Turn float interpolated lons and lats for this day into map coordinates
x, y = map(argofloat.lon_interp(day_range),
argofloat.lat_interp(day_range))
# c. Plot, emphasizing the latest position as a green dot
map.plot(x, y, color='black', zorder=2)
map.scatter(x[-1], y[-1], color='green', s=5, zorder=3)
# d. Label axis
map_axis.set_title(str(wmo_id))
# AXES : par_axes : Parameter vs. Depth for each profile
# 1. Prepare a set of colors to emphasize most recent profile info and grey out old information
cmap = cm.get_cmap('binary')
norm = matplotlib.colors.PowerNorm(vmin=1,
vmax=5,
gamma=3.0,
clip=True)
colors = [
cmap(0.05 + 0.95 * norm(i + 1 - len(day_range) + 5))
for i in range(len(day_range))
]
# 2. Go through the days and...
avg1 = [[] for d in day_range
] #prep to get averages w/ (day, float) indexing
avg2 = [[] for d in day_range
] #prep to get averages w/ (day, float) indexing
for i_d in range(len(day_range)):
# a. Go through each synthetic profile and generate the desired parameters' values
param1_values = []
param2_values = []
for synthprof in self.synth_profiles[day_range[i_d]]:
param1_values.append(
getattr(synthprof, param1 + '_interp')(pres_range))
param2_values.append(
getattr(synthprof, param2 + '_interp')(pres_range))
# b. And iterate through the floats...
for i_f in range(self.n_floats):
# Either plot the values or the delta from average values, and add the average to the list
if delta:
#1. Get the mean values of the parameters first
param1_means = getattr(self.floats[i_f].prof_avg,
param1 + '_interp')(pres_range)
param2_means = getattr(self.floats[i_f].prof_avg,
param2 + '_interp')(pres_range)
#2. Get the differential between this day and the average over time
delta1_values = param1_values[i_f] - param1_means
delta2_values = param2_values[i_f] - param2_means
#3. Append the average to the list of averages
avg1[i_d].append(np.nanmean(delta1_values))
avg2[i_d].append(np.nanmean(delta2_values))
#4. Plot on the corresponding parameter axis
par_axes[i_f].plot(delta1_values[i_f],
param2_values[i_f],
color=colors[i_d])
else:
#1. Add the not-differential average to the list of averages
avg1[i_d].append(np.nanmean(param1_values[i_f]))
avg2[i_d].append(np.nanmean(param2_values[i_f]))
#2. Plot on corresponding parameter axis
par_axes[i_f].plot(param1_values[i_f],
param2_values[i_f],
color=colors[i_d])
# 3. Grid the par_axis plots
for par_axis in par_axes:
par_axis.grid(True)
# 4. Plot the average as a verticle line on the axis
for par_axis, a in zip(par_axes, avg1[-1]):
par_axis.axvline(a, color='red')
# AXIS : day_axis : Parameter Average vs. Time
# 1. Change average indexing from (day, float) to (float, day)
avg1 = np.array(avg1).T
# 2. Plot the average over time for each float
plt.sca(day_axis)
for i in range(self.n_floats):
day_axis.plot(day_range,
avg1[i],
label=self.wmo_ids[i],
linewidth=2)
# 3. Plot the average between floats over time
day_axis.plot(day_range,
np.mean(avg1, axis=0),
label='Average',
linewidth=3,
color='black')
# 4. Add the grid and legend
day_axis.grid(True)
day_axis.legend(loc='lower right')
# 5. Format date on x axis tick labels
xticks = day_axis.get_xticks()
form = [ap.format_jd(xt)
for xt in xticks] # Convert into formatted dates
day_axis.set_xticklabels(form, rotation=90)
# Save and close if desired
if saveas != None:
f.savefig(saveas)
plt.close(f)
else:
plt.show()
def plot_results(allresults,
*,
xy_fn=default_xy_fn,
split_fn=default_split_fn,
group_fn=default_split_fn,
average_group=False,
shaded_std=True,
shaded_err=True,
figsize=None,
legend_outside=False,
resample=0,
smooth_step=1.0,
tiling='vertical',
xlabel=None,
ylabel=None):
'''
Plot multiple Results objects
xy_fn: function Result -> x,y - function that converts results objects into tuple of x and y values.
By default, x is cumsum of episode lengths, and y is episode rewards
split_fn: function Result -> hashable - function that converts results objects into keys to split curves into sub-panels by.
That is, the results r for which split_fn(r) is different will be put on different sub-panels.
By default, the portion of r.dirname between last / and -<digits> is returned. The sub-panels are
stacked vertically in the figure.
group_fn: function Result -> hashable - function that converts results objects into keys to group curves by.
That is, the results r for which group_fn(r) is the same will be put into the same group.
Curves in the same group have the same color (if average_group is False), or averaged over
(if average_group is True). The default value is the same as default value for split_fn
average_group: bool - if True, will average the curves in the same group and plot the mean. Enables resampling
(if resample = 0, will use 512 steps)
shaded_std: bool - if True (default), the shaded region corresponding to standard deviation of the group of curves will be
shown (only applicable if average_group = True)
shaded_err: bool - if True (default), the shaded region corresponding to error in mean estimate of the group of curves
(that is, standard deviation divided by square root of number of curves) will be
shown (only applicable if average_group = True)
figsize: tuple or None - size of the resulting figure (including sub-panels). By default, width is 6 and height is 6 times number of
sub-panels.
legend_outside: bool - if True, will place the legend outside of the sub-panels.
resample: int - if not zero, size of the uniform grid in x direction to resample onto. Resampling is performed via symmetric
EMA smoothing (see the docstring for symmetric_ema).
Default is zero (no resampling). Note that if average_group is True, resampling is necessary; in that case, default
value is 512.
smooth_step: float - when resampling (i.e. when resample > 0 or average_group is True), use this EMA decay parameter (in units of the new grid step).
See docstrings for decay_steps in symmetric_ema or one_sided_ema functions.
'''
if split_fn is None: split_fn = lambda _: ''
if group_fn is None: group_fn = lambda _: ''
sk2r = defaultdict(list) # splitkey2results
for result in allresults:
splitkey = split_fn(result)
sk2r[splitkey].append(result)
assert len(sk2r) > 0
assert isinstance(resample,
int), "0: don't resample. <integer>: that many samples"
if tiling == 'vertical' or tiling is None:
nrows = len(sk2r)
ncols = 1
elif tiling == 'horizontal':
ncols = len(sk2r)
nrows = 1
elif tiling == 'symmetric':
import math
N = len(sk2r)
largest_divisor = 1
for i in range(1, int(math.sqrt(N)) + 1):
if N % i == 0:
largest_divisor = i
ncols = largest_divisor
nrows = N // ncols
figsize = figsize or (6 * ncols, 6 * nrows)
f, axarr = plt.subplots(nrows,
ncols,
sharex=False,
squeeze=False,
figsize=figsize)
groups = list(set(group_fn(result) for result in allresults))
default_samples = 512
if average_group:
resample = resample or default_samples
for (isplit, sk) in enumerate(sorted(sk2r.keys())):
g2l = {}
g2c = defaultdict(int)
sresults = sk2r[sk]
gresults = defaultdict(list)
idx_row = isplit // ncols
idx_col = isplit % ncols
ax = axarr[idx_row][idx_col]
for result in sresults:
group = group_fn(result)
g2c[group] += 1
x, y = xy_fn(result)
if x is None: x = np.arange(len(y))
x, y = map(np.asarray, (x, y))
if average_group:
gresults[group].append((x, y))
else:
if resample:
x, y, counts = symmetric_ema(x,
y,
x[0],
x[-1],
resample,
decay_steps=smooth_step)
l, = ax.plot(x,
y,
color=COLORS[groups.index(group) % len(COLORS)])
g2l[group] = l
if average_group:
for group in sorted(groups):
xys = gresults[group]
if not any(xys):
continue
color = COLORS[groups.index(group) % len(COLORS)]
origxs = [xy[0] for xy in xys]
minxlen = min(map(len, origxs))
def allequal(qs):
return all((q == qs[0]).all() for q in qs[1:])
if resample:
low = max(x[0] for x in origxs)
high = min(x[-1] for x in origxs)
usex = np.linspace(low, high, resample)
ys = []
for (x, y) in xys:
ys.append(
symmetric_ema(x,
y,
low,
high,
resample,
decay_steps=smooth_step)[1])
else:
assert allequal([x[:minxlen] for x in origxs]),\
'If you want to average unevenly sampled data, set resample=<number of samples you want>'
usex = origxs[0]
ys = [xy[1][:minxlen] for xy in xys]
ymean = np.mean(ys, axis=0)
ystd = np.std(ys, axis=0)
ystderr = ystd / np.sqrt(len(ys))
l, = axarr[idx_row][idx_col].plot(usex, ymean, color=color)
g2l[group] = l
if shaded_err:
ax.fill_between(usex,
ymean - ystderr,
ymean + ystderr,
color=color,
alpha=.4)
if shaded_std:
ax.fill_between(usex,
ymean - ystd,
ymean + ystd,
color=color,
alpha=.2)
# https://matplotlib.org/users/legend_guide.html
plt.tight_layout()
if any(g2l.keys()):
ax.legend(g2l.values(),
['%s (%i)' % (g, g2c[g])
for g in g2l] if average_group else g2l.keys(),
loc=2 if legend_outside else None,
bbox_to_anchor=(1, 1) if legend_outside else None)
ax.set_title(sk)
# add xlabels, but only to the bottom row
if xlabel is not None:
for ax in axarr[-1]:
plt.sca(ax)
plt.xlabel(xlabel)
# add ylabels, but only to left column
if ylabel is not None:
for ax in axarr[:, 0]:
plt.sca(ax)
plt.ylabel(ylabel)
return f, axarr
def applyoptions(md, data, options, fig, axgrid, gridindex):
'''
APPLYOPTIONS - apply options to current plot
'plotobj' is the object returned by the specific plot call used to
render the data. This object is used for adding a colorbar.
Usage:
applyoptions(md,data,options)
See also: PLOTMODEL, PARSE_OPTIONS
'''
# get handle to current figure and axes instance
#fig = p.gcf()
ax = axgrid[gridindex]
# {{{ font
fontsize = options.getfieldvalue('fontsize', 8)
fontweight = options.getfieldvalue('fontweight', 'normal')
fontfamily = options.getfieldvalue('fontfamily', 'sans-serif')
font = {
'fontsize': fontsize,
'fontweight': fontweight,
'family': fontfamily
}
# }}}
# {{{ title
if options.exist('title'):
title = options.getfieldvalue('title')
if options.exist('titlefontsize'):
titlefontsize = options.getfieldvalue('titlefontsize')
else:
titlefontsize = fontsize
if options.exist('titlefontweight'):
titlefontweight = options.getfieldvalue('titlefontweight')
else:
titlefontweight = fontweight
#title font
titlefont = font.copy()
titlefont['size'] = titlefontsize
titlefont['weight'] = titlefontweight
ax.set_title(title, **titlefont)
# }}}
# {{{ xlabel, ylabel, zlabel
if options.exist('labelfontsize'):
labelfontsize = options.getfieldvalue('labelfontsize')
else:
labelfontsize = fontsize
if options.exist('labelfontweight'):
labelfontweight = options.getfieldvalue('labelfontweight')
else:
labelfontweight = fontweight
#font dict for labels
labelfont = font.copy()
labelfont['fontsize'] = labelfontsize
labelfont['fontweight'] = labelfontweight
if options.exist('xlabel'):
ax.set_xlabel(options.getfieldvalue('xlabel'), **labelfont)
if options.exist('ylabel'):
ax.set_ylabel(options.getfieldvalue('ylabel'), **labelfont)
if options.exist('zlabel'):
ax.set_zlabel(options.getfieldvalue('zlabel'), **labelfont)
# }}}
# {{{ xticks, yticks, zticks (tick locations)
if options.exist('xticks'):
if options.exist('xticklabels'):
xticklabels = options.getfieldvalue('xticklabels')
ax.set_xticks(options.getfieldvalue('xticks'), xticklabels)
else:
ax.set_xticks(options.getfieldvalue('xticks'))
if options.exist('yticks'):
if options.exist('yticklabels'):
yticklabels = options.getfieldvalue('yticklabels')
ax.set_yticks(options.getfieldvalue('yticks'), yticklabels)
else:
ax.set_yticks(options.getfieldvalue('yticks'))
if options.exist('zticks'):
if options.exist('zticklabels'):
zticklabels = options.getfieldvalue('zticklabels')
ax.set_zticks(options.getfieldvalue('zticks'), zticklabels)
else:
ax.set_zticks(options.getfieldvalue('zticks'))
# }}}
# {{{ xticklabels,yticklabels,zticklabels
if options.getfieldvalue('ticklabels',
'off') == 'off' or options.getfieldvalue(
'ticklabels', 0) == 0:
options.addfielddefault('xticklabels', [])
options.addfielddefault('yticklabels', [])
# TODO check if ax has a z-axis (e.g. is 3D)
if options.exist('xticklabels'):
xticklabels = options.getfieldvalue('xticklabels')
ax.set_xticklabels(xticklabels)
if options.exist('yticklabels'):
yticklabels = options.getfieldvalue('yticklabels')
ax.set_yticklabels(yticklabels)
if options.exist('zticklabels'):
zticklabels = options.getfieldvalue('zticklabels')
ax.set_zticklabels(zticklabels)
# }}}
# {{{ ticklabel notation
#ax.ticklabel_format(style='sci',scilimits=(0,0))
# }}}
# {{{ ticklabelfontsize
if options.exist('ticklabelfontsize'):
for label in ax.get_xticklabels() + ax.get_yticklabels():
label.set_fontsize(options.getfieldvalue('ticklabelfontsize'))
if int(md.mesh.dimension) == 3:
for label in ax.get_zticklabels():
label.set_fontsize(options.getfieldvalue('ticklabelfontsize'))
# }}}
# {{{ view TOFIX
#if int(md.mesh.dimension) == 3 and options.exist('layer'):
# #options.getfieldvalue('view') ?
# ax=fig.gca(projection='3d')
#plt.show()
# }}}
# {{{ axis
if options.exist('axis'):
if options.getfieldvalue('axis', True) == 'off':
ax.ticklabel_format(style='plain')
p.setp(ax.get_xticklabels(), visible=False)
p.setp(ax.get_yticklabels(), visible=False)
# }}}
# {{{ box
if options.exist('box'):
eval(options.getfieldvalue('box'))
# }}}
# {{{ xlim, ylim, zlim
if options.exist('xlim'):
ax.set_xlim(options.getfieldvalue('xlim'))
if options.exist('ylim'):
ax.set_ylim(options.getfieldvalue('ylim'))
if options.exist('zlim'):
ax.set_zlim(options.getfieldvalue('zlim'))
# }}}
# {{{ latlon TODO
# }}}
# {{{ Basinzoom TODO
# }}}
# {{{ ShowBasins TODO
# }}}
# {{{ clim
if options.exist('clim'):
lims = options.getfieldvalue('clim')
assert len(
lims) == 2, 'error, clim should be passed as a list of length 2'
elif options.exist('caxis'):
lims = options.getfieldvalue('caxis')
assert len(
lims) == 2, 'error, caxis should be passed as a list of length 2'
options.addfielddefault('clim', lims)
else:
if len(data) > 0:
lims = [data.min(), data.max()]
else:
lims = [0, 1]
# }}}
# {{{ shading TODO
#if options.exist('shading'):
# }}}
# {{{ grid
if options.exist('grid'):
if 'on' in options.getfieldvalue('grid', 'on'):
ax.grid()
# }}}
# {{{ colormap
if options.exist('colornorm'):
norm = options.getfieldvalue('colornorm')
if options.exist('colormap'):
cmap = options.getfieldvalue('colormap')
cbar_extend = 0
if options.exist('cmap_set_over'):
cbar_extend += 1
if options.exist('cmap_set_under'):
cbar_extend += 2
# }}}
# {{{ contours
if options.exist('contourlevels'):
plot_contour(md, data, options, ax)
# }}}
# {{{ wrapping TODO
# }}}
# {{{ colorbar
if options.getfieldvalue('colorbar', 1) == 1:
if cbar_extend == 0:
extend = 'neither'
elif cbar_extend == 1:
extend = 'max'
elif cbar_extend == 2:
extend = 'min'
elif cbar_extend == 3:
extend = 'both'
cb = mpl.colorbar.ColorbarBase(ax.cax,
cmap=cmap,
norm=norm,
extend=extend)
if options.exist('alpha'):
cb.set_alpha(options.getfieldvalue('alpha'))
if options.exist('colorbarnumticks'):
cb.locator = MaxNLocator(
nbins=options.getfieldvalue('colorbarnumticks', 5))
else:
cb.locator = MaxNLocator(nbins=5) # default 5 ticks
if options.exist('colorbartickspacing'):
locs = np.arange(lims[0], lims[1] + 1,
options.getfieldvalue('colorbartickspacing'))
cb.set_ticks(locs)
if options.exist('colorbarlines'):
locs = np.arange(lims[0], lims[1] + 1,
options.getfieldvalue('colorbarlines'))
cb.add_lines(locs, ['k' for i in range(len(locs))],
np.ones_like(locs))
if options.exist('colorbarlineatvalue'):
locs = options.getfieldvalue('colorbarlineatvalue')
colors = options.getfieldvalue('colorbarlineatvaluecolor',
['k' for i in range(len(locs))])
widths = options.getfieldvalue('colorbarlineatvaluewidth',
np.ones_like(locs))
cb.add_lines(locs, colors, widths)
if options.exist('colorbartitle'):
if options.exist('colorbartitlepad'):
cb.set_label(
options.getfieldvalue('colorbartitle'),
labelpad=options.getfieldvalue('colorbartitlepad'),
fontsize=fontsize)
else:
cb.set_label(options.getfieldvalue('colorbartitle'),
fontsize=fontsize)
cb.ax.tick_params(labelsize=fontsize)
cb.solids.set_rasterized(True)
cb.update_ticks()
cb.set_alpha(1)
cb.draw_all()
if options.exist('colorbarfontsize'):
colorbarfontsize = options.getfieldvalue('colorbarfontsize')
cb.ax.tick_params(labelsize=colorbarfontsize)
# cb.set_ticks([0,-10])
# cb.set_ticklabels([-10,0,10])
if options.exist('colorbarticks'):
colorbarticks = options.getfieldvalue('colorbarticks')
cb.set_ticks(colorbarticks)
plt.sca(ax) # return to original axes control
# }}}
# {{{ expdisp
if options.exist('expdisp'):
expdisp(ax, options)
# }}}
# {{{ area TODO
# }}}
# {{{ text
if options.exist('text'):
text = options.getfieldvalue('text')
textx = options.getfieldvalue('textx')
texty = options.getfieldvalue('texty')
textcolor = options.getfieldvalue('textcolor')
textweight = options.getfieldvalue('textweight')
textrotation = options.getfieldvalue('textrotation')
textfontsize = options.getfieldvalue('textfontsize')
for label, x, y, size, color, weight, rotation in zip(
text, textx, texty, textfontsize, textcolor, textweight,
textrotation):
ax.text(x,
y,
label,
transform=ax.transAxes,
fontsize=size,
color=color,
weight=weight,
rotation=rotation)
# }}}
# {{{ north arrow TODO
# }}}
# {{{ scale ruler TODO
# }}}
# {{{ streamlines TOFIX
if options.exist('streamlines'):
plot_streamlines(md, options, ax)
def plotXsection(startpoints,
endpoints,
names,
geo_model,
lith,
surflith,
vscale=1,
unitnames=None):
'''Plots an approximate cross-section between the two specified points, using an elevation-cropped array (cells above the land surface should have nan values). Does not work well for N-S or nearly N-S xsections - use gempy built-in for that.
startpoints: [[x1,y1],[x2,y2],...] float or [[col1,row1],[col1,row2],...] integer array of coordinates of starting points A
endpoints: [[x1,y1],[x2,y2],...] float or [[col1,row1],[col1,row2],...] integer array of coordinates of ending points B
names: [['A','B'],['C','D'],...] string array of names for starting and ending points
grid_info: [xmin,xmax,xres,dx,ymin,ymax,yres,dy,zmin,zmax,zres,dz] model grid info (can get these using importDEM() function if model grid is same as DEM grid)
lith: elevation-cropped array of lithologic unit indices of dimensions (nrow,ncol,nlay), i.e. (yres,xres,zres). Can use uncropped array, but will plot above land surface.
surflith: array of lithologic unit values at the land surface, of dimensions (yres,xres)
vscale: vertical exaggeration factor (y/x, defaults to 1)
colors: dictionary of color & unit names to use (in order from youngest to oldest) OR 'gempy' to use default gempy colormap
'''
#Get coordinate info from grid & create VTK cells info:
xmin = geo_model.grid.regular_grid.extent[0] #min coordinate value (left)
xmax = geo_model.grid.regular_grid.extent[1] #max coordinate value (right)
xres = geo_model.grid.regular_grid.resolution[0] #number of pixels
dx = (xmax - xmin) / xres #pixel width
ymin = geo_model.grid.regular_grid.extent[2]
ymax = geo_model.grid.regular_grid.extent[3]
yres = geo_model.grid.regular_grid.resolution[1]
dy = (ymax - ymin) / yres
zmin = geo_model.grid.regular_grid.extent[4]
zmax = geo_model.grid.regular_grid.extent[5]
zres = geo_model.grid.regular_grid.resolution[2]
dz = (zmax - zmin) / zres
#Generate colormap:
cmap = matplotlib.colors.ListedColormap(
geo_model.surfaces.colors.colordict.values()) #set colormap
norm = matplotlib.colors.Normalize(
vmin=1,
vmax=len(geo_model.surfaces.colors.colordict)) #set normalization
#Plot geologic map once for reference:
f1, ax1 = plt.subplots(1, 1, figsize=(10, 10)) #create empty figure
plt.imshow(surflith, cmap=cmap,
norm=norm) #plot geology (normalized to gempy color range)
f2, ax2 = plt.subplots(len(startpoints), 1, figsize=(
15,
20)) #create figure and axes objects for subplots (one per xsection)
for i in range(len(startpoints)): #loop over number of sections
#Get starting coordinates:
xA = startpoints[i][0] #get starting x coordinate
yA = startpoints[i][1] #get starting y coordinate
xB = endpoints[i][0] #get ending x coordinate
yB = endpoints[i][1] #get ending y coordinate
#Calculate corresponding row,col
if type(
xA
) != int: #if coordinates are NOT integers (i.e. not row,col numbers), convert them
colA = (xA -
xmin) // dx #col:x calculate column index c = (x1-x0)/dx
rowA = yres - (
(yA - ymin) // dy
) #row:y calculate row index r = ymax - (y1-y0)/dy
colB = (xB - xmin) // dx
rowB = yres - ((yB - ymin) // dy)
else: #if coordinates are already in row,col format
colA = xA
rowA = yA
colB = xB
rowB = yB
#Calculate line equation between points A and B:
m = (rowB - rowA) / (colB - colA
) #calculate slope m = (y2-y1)/(x2-x1)
b = -m * colA + rowA #calculate intercept b = m*x1 + y1 (slope is neg here bc y axis is flipped)
#Calculate true distance (not # of cells) between points A and B:
#distance = ((xB-xA)**2 + (yB-yA)**2)**.5
#xsizes.append(distance*.001)
#Get xy indices for cells intersected by the x-sec line, then get z values for those xy points:
xvals = np.arange(
colA, colB) #generate array of x values between the two points
xvals = xvals.astype(int) #convert to integer
yvals = m * xvals + b #calculate corresponding y values y = mx + b
yvals = yvals.astype(int) #convert to integers to be able to slice
#xsec = lith[yvals,xvals,:].T #select x-sec to plot and transpose to make it plot horizontally
xsec = lith[
xvals,
yvals, :].T #select x-sec to plot and transpose to make it plot horizontally
#Plotting:
#Add xsection lines to geologic map:
plt.figure(f1.number) #make the map the active figure
plt.plot([colA, colB], [rowA, rowB], 'k') #plot x-sec location line
plt.annotate(names[i][0], xy=(colA, rowA),
xytext=(colA - 4, rowA + 4)) #annotate start point
plt.annotate(names[i][1], xy=(colB, rowB),
xytext=(colB + 1, rowB - 1)) #annotate start point
plt.ylim(bottom=yres, top=0)
plt.xlim(left=0, right=xres)
#Plot cross-sections in a new figure:
#Set and get correct subplot axes:
if len(
startpoints
) == 1: #check if there are more than 1 subplots (for indexing purposes)
plt.sca(
ax2
) #make current subplot axes active (automatically makes fig active too)
cax = plt.gca() #get current axes object
else:
plt.sca(ax2[i])
cax = plt.gca()
cax.imshow(xsec, origin="lower", cmap=cmap,
norm=norm) #plot (with down=lower z indices)
#cax.imshow(xsec, origin="lower", cmap=cmap) #plot (with down=lower z indices)
cax.set_aspect(vscale * dz / dx) #apply vertical exaggeration
cax.set_ylim(bottom=0, top=zres) #set y limit to zres
cax.set_title(names[i][0] + names[i][1])
cax.set_anchor('W') #align left (West)
#Set ticks to accurately reflect elevation (masl):
locs = cax.get_yticks() #get tick locations
nlabels = len(cax.get_yticklabels()) #get number of initial ticks
labels = np.linspace(zmin, zmax,
nlabels) #generate list of tick labels
ticks = cax.set(yticks=locs,
yticklabels=labels) #set tick locations and labels
return f1, ax1, f2, ax2
def colorbar_to_axis(self, toaxis, clim, cmap, vmin=None, under='gray',
vmax=None, over='red', title=None, ycb=10.,
fz_title=15, ticks='complete', fz_ticks=10,
outline=False, orientation='vertical'):
"""Add a colorbar to a particular axis.
Parameters
----------
toaxis : int
Integer to specify the axis to which add the colorbar.
clim : tuple/list
A tuple of float/int describing the limit of the colorbar.
cmap : str
The colormap to use ('viridis', 'jet', 'Spectral'...)
vmin : float | None
Minimum threshold. See the under parameter for further details.
under : string/tuple/list | 'gray'
Every values bellow vmin will be set to the under color.
vmax : float | None
Maximum threshold. See the over parameter for further details.
over : string/tuple/list | 'red'
Every values over vmax will be set to the over color.
title : string | None
Title of the colorbar.
ycb : float | 10.
Distance between the colorbar and it title (if defined).
fz_title : int/float | 15
The fontsize of colorbar title.
ticks : string/int/float/np.ndarray | 'complete'
Ticks of the colorbar. This parameter is only active if clim
is defined. Use 'complete' to see only the minimum,
maximum, vmin and vmax (if defined). Use 'minmax' to only see
the maximum and minimum. If ticks is a float, an linear
interpolation between the maximum and minimum will be used.
Finally, if ticks is a NumPy array, it will be used as colorbar
ticks directly.
fz_ticks : int/float | 10
Fontsize of tick labels.
outline : bool | False
Specify if the box arround the colorbar have to be displayed.
orientation : string | 'vertical'
Colorbar orientation. Use either 'vertical' or 'horizontal'.
See also
--------
shared_colorbar : colorbar shared by multiple subplots.
"""
# ----------------- CHECKING -----------------
if orientation not in ['vertical', 'horizontal']:
raise ValueError("Colorbar orientation must either be 'vertical' "
"or 'horizontal'")
# ----------------- CURRENT AXIS -----------------
cim = self._im[toaxis]
plt.sca(self._ax[toaxis])
# ----------------- COLORMAP -----------------
# Manually create the colormap :
cmap = self._customcmap(cmap, clim, vmin, under, vmax, over)
cim.set_cmap(cmap)
# ----------------- COLORBAR -----------------
cb = plt.colorbar(cim, shrink=0.7, pad=0.01, aspect=10,
orientation=orientation)
# # ----------------- CLIM -----------------
cim.set_clim(*clim)
cb.set_clim(*clim)
# ----------------- CBAR -----------------
self._cbar(cb, cmap, clim, vmin, under, vmax, over, title, ycb,
fz_title, ticks, fz_ticks, outline, orientation, self._tcol)
