def setup_axes(myg, num):
""" create a grid of axes whose layout depends on the aspect ratio of the
domain """
L_x = myg.xmax - myg.xmin
L_y = myg.ymax - myg.ymin
f = plt.figure(1)
cbar_title = False
if L_x > 2 * L_y:
# we want num rows:
axes = AxesGrid(f,
111,
nrows_ncols=(num, 1),
share_all=True,
cbar_mode="each",
cbar_location="top",
cbar_pad="10%",
cbar_size="25%",
axes_pad=(0.25, 0.65),
add_all=True,
label_mode="L")
cbar_title = True
elif L_y > 2 * L_x:
# we want num columns: rho |U| p e
axes = AxesGrid(f,
111,
nrows_ncols=(1, num),
share_all=True,
cbar_mode="each",
cbar_location="right",
cbar_pad="10%",
cbar_size="25%",
axes_pad=(0.65, 0.25),
add_all=True,
label_mode="L")
else:
# 2-d grid of plots
ny = int(math.sqrt(num))
nx = num // ny
axes = AxesGrid(f,
111,
nrows_ncols=(nx, ny),
share_all=True,
cbar_mode="each",
cbar_location="right",
cbar_pad="2%",
axes_pad=(0.65, 0.25),
add_all=True,
label_mode="L")
return f, axes, cbar_title
def demo_grid_with_single_cbar_log(fig):
"""
A grid of 2x2 images with a single colorbar and log scaling
"""
grid = AxesGrid(
fig,
111, # modified to be only subplot
nrows_ncols=(2, 2),
axes_pad=0.0,
share_all=True,
label_mode="L",
cbar_location="top",
cbar_mode="single",
)
Z, extent = get_demo_image()
Z -= np.min(Z) # modified to make data positive
for i in range(4):
im = grid[i].imshow(Z,
extent=extent,
interpolation="nearest",
norm=LogNorm()) # modified to log-scale display
#plt.colorbar(im, cax = grid.cbar_axes[0])
grid.cbar_axes[0].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(False)
# This affects all axes as share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
plt.show()
def img_pro(path):
F = plt.figure(1, (15,20))
grid = AxesGrid(F, 111, nrows_ncols=(4,4), axes_pad=0, label_mode='1')
for i in range(16):
char = map_characters[i]
list = [path+k for k in os.listdir(path) if char in k]
image = cv2.imread(np.random.choice(list))
img = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
pic = cv2.resize(image, (pic_size,pic_size)).astype('float32')/255
a = model.predict(pic.reshape(1, pic_size, pic_size, 3))[0]
actual = char.split('_')[0].title()
text = sorted(['{:s} : {:.1f}%'.format(map_characters[i].split('_')[0].title(), 100*v) for k,v in enumerate(a)],
key=lambda x:float(x.split(':')[1].split('%')[0]), reverse=True)[:3]
img = cv2.resize(img, (352, 352))
cv2.rectangle(img, (0,260), (215,352), (255,255,255), -1)
font = cv2.FONT_HERSHEY_SIMPLEX
cv2.putText(img, 'Actual : %s' % actual, (10,280), font, 0.7, (0,0,0), 2, cv2.LINE_AA)
for k,t in enumerate(text):
cv2.putText(img, t, (10, 300+k*18), font, 0.65, (0,0,0), 2, cv2.LINE_AA)
grid[i].imshow(img)
plt.show()
def demo_right_cbar(fig):
"""
A grid of 2x2 images. Each row has its own colorbar.
"""
grid = AxesGrid(
fig,
122, # similar to subplot(122)
nrows_ncols=(2, 2),
axes_pad=0.10,
label_mode="1",
share_all=True,
cbar_location="right",
cbar_mode="edge",
cbar_size="7%",
cbar_pad="2%",
)
Z, extent = get_demo_image()
cmaps = [plt.get_cmap("spring"), plt.get_cmap("winter")]
for i in range(4):
im = grid[i].imshow(Z,
extent=extent,
interpolation="nearest",
cmap=cmaps[i // 2])
if i % 2:
grid.cbar_axes[i // 2].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(True)
cax.axis[cax.orientation].set_label('Foo')
# This affects all axes because we set share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
def plot_filter(image_path, layer_name, output_dir):
base_model = VGG16(weights='imagenet')
x = load_images([image_path])
model = Model(input=base_model.input,
output=base_model.get_layer(layer_name).output)
layer_output = model.predict(x)
side = int(layer_output.shape[-1]**0.5)
fig = plt.figure()
grid = AxesGrid(fig,
111,
nrows_ncols=(side, side),
axes_pad=0.0,
share_all=True)
for i in range(side**2):
im = grid[i].imshow(layer_output[0, :, :, i], interpolation="nearest")
grid.cbar_axes[0].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(False)
for ax in grid.axes_all:
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
output_dir = Path(output_dir)
fig_file = '{}-{}.pdf'.format(Path(image_path).stem, layer_name)
plt.savefig(str(output_dir / fig_file))
def draw_grid(fig, data_array, x_min, x_max, y_min, y_max):
"""
A grid of 2x2 images with a single colorbar
"""
grid = AxesGrid(fig, 111, # similar to subplot(142)
nrows_ncols=(1, 1),
axes_pad=0.2,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
Z = data_array
extent = (x_min, x_max, y_min, y_max)
im = grid[0].imshow(Z, extent=extent, interpolation="nearest")
grid[0].set_aspect(2.0, adjustable='box')
grid.cbar_axes[0].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(True)
# This affects all axes as share_all = True.
grid.axes_llc.set_xticks(range(x_min, x_max, 10)) # Set x-axes sequence
grid.axes_llc.set_yticks(range(y_min, y_max, 10))
def plot(D, i, flux_vol, xlabel, ylabel, name, sigma):
flux = flux_vol.sum(axis=i)
flux = nd.gaussian_filter(flux, sigma=(sigma, sigma), order=0)
fig = plt.figure(figsize=(6.5, 6.0))
ax = AxesGrid(
fig,
111, # similar to subplot(122)
nrows_ncols=(1, 1),
axes_pad=0.0,
label_mode="1",
share_all=True,
cbar_location="right",
cbar_mode="edge",
cbar_size="2%",
cbar_pad="0%",
)
neg = ax[0].imshow(flux.T,
origin='lower',
extent=[D.x2[0], D.x2[-1], D.x3[0], D.x3[-1]])
ax.cbar_axes[0].colorbar(neg)
# fig.subplots_adjust(top=0.99,bottom=0.11,left=0.11,right=0.99)
ax.axes_llc.set_xlabel(xlabel, fontsize=20)
ax.axes_llc.set_ylabel(ylabel, fontsize=20)
fig.savefig(name)
def triple_plot(ds,fields,lfpps={},**kwargs):
fig = plt.figure(figsize=(15,5))
grid = AxesGrid(fig, (0.075,0.075,0.85,0.85),
nrows_ncols = (1, 3),
axes_pad = 0.9,
#label_mode = "1",
share_all = True,
cbar_location="right",
cbar_mode="each",
cbar_size="3%",
cbar_pad="0%")
#fields = [("flash",'soundspeed'), ("flash",'velocityz'), ("flash",'temperature'), ("flash",'density')]
ps = []
for field in fields:
fpp = lfpps.get(field,{})
if not(fpp):
fpp = fpps.get(field,{})
ps.append(slice_plot(ds,field,plotprops=fpp,**kwargs))
for i, field in enumerate(fields):
plot = ps[i].plots[field]
plot.figure = fig
plot.axes = grid[i].axes
plot.cax = grid.cbar_axes[i]
ps[i]._setup_plots()
return fig
def plot_filters_single_channel(t):
nrows=25
ncols=25
fig,ax = plt.subplots(nrows=25, ncols=25, figsize = (38,38))
count = 0
#looping through all the kernels in each channel
grid = AxesGrid(fig, 111,
nrows_ncols=(25, 25),
axes_pad=0.05,
cbar_mode='single',
cbar_location='right',
cbar_pad=0.1
)
count = 0
vmin = np.amin(t)
vmax = np.amax(t)
ax = []
for x in grid:
ax.append(x)
count = 0
#looping through all the kernels in each channel
for i in range(1, 26):
for j in range(1, 26):
pcm = ax[count].imshow(t[i-1, j-1],vmin = vmin, vmax = vmax, cmap = "plasma")
ax[count].axis('off')
cbar = ax[count].cax.colorbar(pcm)
count+=1
cbar = grid.cbar_axes[0].colorbar(pcm)
plt.savefig("w_gate.png")
return
def _create_figure(self, figsize=None, **kwargs):
# creates a figure and then an AxesGrid in the 111 position
fig = Figure(figsize=figsize)
grid = AxesGrid(fig, 111, **kwargs)
self.grid = grid
self.next_grid = {grid: 0}
return fig, grid
def plotting_weights(save_folder, filename, mat, true_states=None, estimated_states=None, soz_ch_ids=None, sel_win_num=None):
title_fontSize = 40
# ictal_indices = np.argwhere(true_states!=0).astype(np.int)
# ictal_indices = ictal_indices.reshape((ictal_indices.size,))
num_wind = len(mat)
plot_num_rows = int(np.ceil(num_wind**0.5))
plot_num_cols = int(np.ceil(num_wind/plot_num_rows))
if not os.path.exists(save_folder):
os.makedirs(save_folder)
fig = plt.figure(num=None, figsize=(60, 40), dpi=120)
grid = AxesGrid(fig, 111,
nrows_ncols=(plot_num_rows, plot_num_cols),
axes_pad=1,
cbar_mode='single',
cbar_location='right',
cbar_pad=0.1)
i = 0
for ax in grid:
if(i>=len(mat)):
break
N = int(mat[i].size**0.5)
im = ax.imshow(np.reshape(mat[i], (N, N)))
if(true_states is not None and true_states[i] != 0 and estimated_states is not None):
ax.set_title('--'+ str(estimated_states[i])+'--', fontsize=title_fontSize+5)
elif(true_states is not None and true_states[i] == 0 and estimated_states is not None):
ax.set_title(str(estimated_states[i]), fontsize=title_fontSize)
i += 1
cbar = ax.cax.colorbar(im)
cbar.ax.tick_params(labelsize=30)
cbar = grid.cbar_axes[0].colorbar(im)
plt.savefig(save_folder + 'W_' + filename + '.png')
def demo_grid_with_each_cbar(fig):
"""
A grid of 2x2 images. Each image has its own colorbar.
"""
grid = AxesGrid(
F,
143, # similar to subplot(143)
nrows_ncols=(2, 2),
axes_pad=0.1,
label_mode="1",
share_all=True,
cbar_location="top",
cbar_mode="each",
cbar_size="7%",
cbar_pad="2%",
)
Z, extent = get_demo_image()
for i in range(4):
im = grid[i].imshow(Z, extent=extent, interpolation="nearest")
grid.cbar_axes[i].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(False)
# This affects all axes because we set share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
def double_plot(settings, ds):
fig = plt.figure()
grid = AxesGrid(fig, (0.09,0.09,0.8,0.8),
nrows_ncols = (1, 2),
axes_pad = 0.05,
label_mode = "L",
share_all = True,
cbar_location="right",
cbar_mode="single",
cbar_size="5%",
cbar_pad="1.2%")
for i, fn in enumerate(fns):
if settings["option"] == "top_down":
slc = yt.SlicePlot(ds,
'z' ,
settings["field"],
center=settings["center"],
width=settings["width"],
fontsize=settings["font"])
if settings["streamlines"]:
slc.annotate_streamlines('velocity_x',
'velocity_y',
density=1.5,
factor=16,
plot_args={
'color': 'black',
'linewidth': 0.25})
if settings["option"] == "side_on":
slc = yt.OffAxisSlicePlot(ds,
settings["L"],
settings["field"],
center=settings["center"],
north_vector=settings["north_vector"],
width=settings["width"],
fontsize=settings["font"])
slc.set_xlabel('x $\ (\mathrm{R}_{\odot})$')
slc.set_ylabel('z $\ (\mathrm{R}_{\odot})$')
if settings["streamlines"]:
slc.annotate_streamlines('magnetic_field_x',
'magnetic_field_z',
density=1.5,
factor=16,
plot_args={
'color': 'white',
'linewidth': 0.75})
slc.set_cmap(field=settings["field"], cmap='jet')
slc.set_zlim(settings["field"], settings["lim"][0], settings["lim"][1])
plot = slc.plots[settings["field"]]
plot.figure = fig
plot.axes = grid[i].axes
plot.cax = grid.cbar_axes[i]
slc._setup_plots()
slc.save("plots/"+settings["save_name"]+".pdf")
def AX_perturbations(fig, perturbations):
"""
A grid of 1x10 images representing the perturbations of the
generated AXs - with a single coolwarm colorbar at the right
"""
grid = AxesGrid(
fig,
211, # similar to subplot(211)
nrows_ncols=(1, 10),
axes_pad=0.0,
share_all=True,
label_mode="1",
cbar_location="right",
cbar_mode="single",
cbar_size="7%",
cbar_pad="3%",
)
for i in range(nb_classes):
img = perturbations[i].reshape([28, 28])
im = grid[i].imshow(img,
interpolation="nearest",
cmap=cm.coolwarm,
vmin=-1.,
vmax=1.)
grid[i].tick_params(which='both',
bottom='off',
left='off',
labelbottom='off',
labelleft='off')
import matplotlib as mpl
norm_ = mpl.colors.Normalize(vmin=-1., vmax=1.)
grid.cbar_axes[0].colorbar(im, norm=norm_)
#grid.cbar_axes[0].set_yticklabels(['-1', '0', '1'])
grid.cbar_axes[0].set_yticks((-1, 0, 1))
def AX_actual(fig, adv_x, top_1, confidence, ylabel):
"""
A grid of 1x10 images displaying the actual AXs along with
their predicted labels and confidences"""
grid = AxesGrid(
fig,
111, # similar to subplot(212)
nrows_ncols=(1, 10),
axes_pad=0.0,
share_all=True,
label_mode="all")
for i in range(nb_classes):
img = adv_x[i].reshape([28, 28])
im = grid[i].imshow(img, cmap='gray')
grid[i].tick_params(which='both',
bottom='off',
left='off',
labelbottom='off',
labelleft='off')
conf = str(
Decimal(str(confidence[i])).quantize(Decimal('0.01'),
rounding=ROUND_DOWN))
#xlabel = str(top_1[i]) + " " + "(" + '{0:.2f}'.format(confidence[i]) + ')'
xlabel = str(top_1[i]) + " " + "(" + conf + ")"
grid[i].set_xlabel(xlabel, labelpad=2.0, fontsize=12)
grid[9].yaxis.set_label_position("right")
# ylabel should be a string
grid[9].set_ylabel(ylabel, labelpad=14.0, fontsize=15, rotation=270)
def plot_multiple_model_weights(weights_to_plot):
"""
Plot the weights of different models side by side as a heat maps.
"""
models_weights = np.array(
[rectangularfy(weights) for weights in weights_to_plot])
vmax = models_weights.max()
vmin = models_weights.min()
fig = plt.figure()
grid = AxesGrid(
fig,
111,
nrows_ncols=(1, len(models_weights)),
axes_pad=0.05,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
for axis, model_weights in zip(grid, models_weights):
heatmap = axis.imshow(model_weights,
cmap="coolwarm",
vmax=vmax,
vmin=vmin)
axis.set_xticks([])
axis.set_yticks([])
grid.cbar_axes[0].colorbar(heatmap)
plt.show()
def displayMaps(self, map_list, figFilename, n_rows=2):
print 'map_list: ', (np.array(map_list)).shape
fig = plt.figure(figsize=(200, 200))
grid = AxesGrid(
fig,
111,
nrows_ncols=(n_rows, int(np.round(len(map_list) / n_rows))),
axes_pad=0.02,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="each",
)
vmin = float('inf')
vmax = -float('inf')
for cur_map in map_list:
cur_map_min = np.min(cur_map)
cur_map_max = np.max(cur_map)
vmin = min(vmin, cur_map_min)
vmax = max(vmax, cur_map_max)
print 'DEBUG:vmax, vmin: ', vmax, vmin
for cur_map, ax in zip(map_list, grid):
im = ax.imshow(cur_map, vmin=vmin, vmax=vmax)
#im = ax.imshow(cur_map, vmin=self.noise_floor_dB, vmax=0.0)
grid.cbar_axes[0].colorbar(im)
#plt.show(block = False)
#plt.colorbar()
plt.savefig(figFilename)
def demo_grid_with_each_cbar_labelled(fig):
"""
A grid of 2x2 images. Each image has its own colorbar.
"""
grid = AxesGrid(
F,
144, # similar to subplot(144)
nrows_ncols=(2, 2),
axes_pad=(0.45, 0.15),
label_mode="1",
share_all=True,
cbar_location="right",
cbar_mode="each",
cbar_size="7%",
cbar_pad="2%",
)
Z, extent = get_demo_image()
# Use a different colorbar range every time
limits = ((0, 1), (-2, 2), (-1.7, 1.4), (-1.5, 1))
for i in range(4):
im = grid[i].imshow(Z,
extent=extent,
interpolation="nearest",
vmin=limits[i][0],
vmax=limits[i][1])
grid.cbar_axes[i].colorbar(im)
for i, cax in enumerate(grid.cbar_axes):
cax.set_yticks((limits[i][0], limits[i][1]))
# This affects all axes because we set share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
def quad_plot(ds,fields,width=(5000,"km")):
fig = plt.figure()
grid = AxesGrid(fig, (0.075,0.075,0.85,0.85),
nrows_ncols = (2, 2),
axes_pad = 0.9,
#label_mode = "1",
share_all = True,
cbar_location="right",
cbar_mode="each",
cbar_size="3%",
cbar_pad="0%")
#fields = [("flash",'soundspeed'), ("flash",'velocityz'), ("flash",'temperature'), ("flash",'density')]
ps = []
for field in fields:
ps.append(slice_plot(ds,field,width=width))
for i, field in enumerate(fields):
plot = ps[i].plots[field]
plot.figure = fig
plot.axes = grid[i].axes
plot.cax = grid.cbar_axes[i]
ps[i]._setup_plots()
return fig
def multiplot():
field = 'density'
view = 'projection'
field = 'O_p5_number_density'
track = Table.read('complete_track', format='ascii')
track.sort('col1')
outs = [x+55 for x in range(400)]
for n in outs:
fig = plt.figure()
grid = AxesGrid(fig, (0.5,0.5,1.5,1.5),
nrows_ncols = (1, 5),
axes_pad = 0.1,
label_mode = "1",
share_all = True,
cbar_location="right",
cbar_mode="edge",
cbar_size="5%",
cbar_pad="0%")
strset = 'DD00'+str(n)
if (n > 99): strset = 'DD0'+str(n)
fields = [field, field, field, field, field]
snaps = ['nref10_track_2/'+strset+'/'+strset, 'nref10_track_lowfdbk_1/'+strset+'/'+strset,
'nref10_track_lowfdbk_2/'+strset+'/'+strset, 'nref10_track_lowfdbk_3/'+strset+'/'+strset,
'nref10_track_lowfdbk_4/'+strset+'/'+strset]
for i, (field, snap) in enumerate(zip(fields, snaps)):
ds = yt.load(snap)
zsnap = ds.get_parameter('CosmologyCurrentRedshift')
trident.add_ion_fields(ds, ions=['C IV', 'O VI','H I','Si III'])
centerx = np.interp(zsnap, track['col1'], 0.5*(track['col2']+track['col5']) )
centery = np.interp(zsnap, track['col1'], track['col3']+30./143886.)
centerz = np.interp(zsnap, track['col1'], 0.5*(track['col4']+track['col7']) )
center = [centerx, centery, centerz]
box = ds.r[ center[0]-200./143886:center[0]+200./143886, center[1]-250./143886.:center[1]+250./143886., center[2]-40./143886.:center[2]+40./143886.]
# projection
p = yt.ProjectionPlot(ds, 'z', field, center=center, width=((120,'kpc'),(240,'kpc')), data_source=box)
if (field == 'density'):
p.set_unit('density', 'Msun / pc**2')
p.set_zlim('density', 0.01, 1000)
if ('O_p5' in field):
p.set_zlim("O_p5_number_density",1e11,1e15)
if (i < 1): p.annotate_timestamp(corner='upper_left', redshift=True, draw_inset_box=True, text_args={'color':'white', 'size':'small'} )
# This forces the ProjectionPlot to redraw itself on the AxesGrid axes.
plot = p.plots[field]
plot.figure = fig
plot.axes = grid[i].axes
p._setup_plots() # Finally, redraw the plot.
plt.savefig(strset+'_multiplot_'+field+'_projection.png', bbox_inches='tight')
def demo_grid_with_single_cbar(fig):
"""
A grid of 2x2 images with a single colorbar
"""
grid = AxesGrid(
fig,
142, # similar to subplot(142)
nrows_ncols=(2, 2),
axes_pad=0.0,
share_all=True,
label_mode="L",
cbar_location="top",
cbar_mode="single",
)
Z, extent = get_demo_image()
for i in range(4):
im = grid[i].imshow(Z, extent=extent, interpolation="nearest")
#plt.colorbar(im, cax = grid.cbar_axes[0])
grid.cbar_axes[0].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(False)
# This affects all axes as share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
def displayMaps(self, map_list, figFilename=None, n_rows=1, title=''):
fig = plt.figure()
grid = AxesGrid(
fig,
111,
nrows_ncols=(n_rows, int(np.round(len(map_list) / n_rows))),
axes_pad=0.01,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
vmin = float('inf')
vmax = -float('inf')
for cur_map in map_list:
cur_map_min = np.min(cur_map)
cur_map_max = np.max(cur_map)
vmin = min(vmin, cur_map_min)
vmax = max(vmax, cur_map_max)
#print 'DEBUG:vmax, vmin: ',vmax, vmin
for cur_map, ax in zip(map_list, grid):
im = ax.imshow(cur_map, vmin=vmin, vmax=vmax)
grid.cbar_axes[0].colorbar(im)
if figFilename is not None:
plt.savefig(figFilename)
else:
plt.show(block=False)
def heatmap_in_one_figure(vals, pars, cmap=None):
"""
vals: lists of dataframes
pars: nrow and ncol of plots in the figure
"""
from mpl_toolkits.axes_grid1 import AxesGrid
fig = plt.figure()
grid = AxesGrid(
fig,
111,
nrows_ncols=pars,
axes_pad=0.05,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
for val, ax in zip(vals, grid):
myplot3d = plot3D(val)
hm = ax.pcolor(myplot3d.x,
myplot3d.y,
myplot3d.z,
vmin=0,
vmax=1,
cmap=cmap)
grid.cbar_axes[0].colorbar(hm)
return fig, ax
def demo_bottom_cbar(fig):
"""
A grid of 2x2 images with a colorbar for each column.
"""
grid = AxesGrid(
fig,
121, # similar to subplot(121)
nrows_ncols=(2, 2),
axes_pad=0.10,
share_all=True,
label_mode="1",
cbar_location="bottom",
cbar_mode="edge",
cbar_pad=0.25,
cbar_size="15%",
direction="column")
Z, extent = get_demo_image()
cmaps = [plt.get_cmap("autumn"), plt.get_cmap("summer")]
for i in range(4):
im = grid[i].imshow(Z,
extent=extent,
interpolation="nearest",
cmap=cmaps[i // 2])
if i % 2:
grid.cbar_axes[i // 2].colorbar(im)
for cax in grid.cbar_axes:
cax.toggle_label(True)
cax.axis[cax.orientation].set_label("Bar")
# This affects all axes as share_all = True.
grid.axes_llc.set_xticks([-2, 0, 2])
grid.axes_llc.set_yticks([-2, 0, 2])
def displayMaps(self, map_list, figFilename, n_rows = 1):
fig = plt.figure()
fig.suptitle("Maximum %-diff in input map-pairs", fontsize=16)
nrows_ncols = (n_rows, int( np.round( len(map_list)/n_rows) ))
grid = AxesGrid(fig, 111,
nrows_ncols= nrows_ncols,
axes_pad=0.1,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
vmin = float('inf')
vmax = -float('inf')
for cur_map in map_list:
cur_map_min = np.min(cur_map)
cur_map_max = np.max(cur_map)
vmin = min(vmin, cur_map_min)
vmax = max(vmax, cur_map_max)
if (vmax == vmin):
vmax += 1
for cur_map, ax in zip(map_list, grid):
im = ax.imshow(cur_map, vmin=vmin, vmax=vmax)
grid.cbar_axes[0].colorbar(im)
#plt.show(block = False)
plt.savefig(figFilename)
def create_maps(self):
g = self.chan_map.geometry
# this is an over-estimate of inches per pixel, maybe
# find another rule
inch_per_pixel = 0.039 * self.mm_per_pixel
# size of array map: width, height
img_size = g[1] * inch_per_pixel, g[0] * inch_per_pixel
figsize = self.map_col * img_size[0], self.map_row * img_size[1]
figwin = SimpleFigure(figsize=figsize)
fig = figwin.figure
text_size_inch = 11 / fig.dpi
grid = AxesGrid(
fig, 111, nrows_ncols=(self.map_row, self.map_col),
axes_pad=1.5 * text_size_inch,
cbar_mode='single', cbar_location='right',
cbar_pad='2%', cbar_size='4%'
)
for ax in grid:
ax.axis('off')
self._grid = grid
self._g_idx = 0
self._cbar = None
# hold onto this or it disappears
figwin.show()
self._figwin = figwin
def plotImgMosaic(imgOrig, extent, origin='upper', T=0):
fig = plt.figure()
grid = AxesGrid(
fig,
111, # similar to subplot(111)
nrows_ncols=(3, 2),
axes_pad=0.0,
share_all=True,
label_mode="all",
cbar_location="top",
cbar_mode="single",
)
img = dict(imgOrig)
for im in img:
img[im] = np.squeeze(img[im])
if T:
img[im] = img[im].T
ij = "21"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[0].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[0].set_ylabel(ij)
ij = "41"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[1].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[1].set_ylabel(ij)
grid[1].yaxis.set_ticks_position('right')
grid[1].yaxis.set_label_position('right')
ij = "31"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[2].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[2].set_ylabel(ij)
ij = "42"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[3].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[3].set_ylabel(ij)
grid[3].yaxis.set_ticks_position('right')
grid[3].yaxis.set_label_position('right')
ij = "32"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[4].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[4].set_ylabel(ij)
ij = "43"
toPlot = 20 * np.log10(np.abs(img[ij]))
im = grid[5].imshow(toPlot, cmap='jet', extent=extent, origin=origin)
grid[5].set_ylabel(ij)
grid[5].yaxis.set_ticks_position('right')
grid[5].yaxis.set_label_position('right')
grid.cbar_axes[0].colorbar(im)
return fig, grid
def plot_activation_gradients(all_timesteps_activations: np.array,
neuron_heatmap_size: tuple,
show_title=True,
absolute=True,
save=None):
"""
Plot the changes in activation values between time steps as heat maps for one single sample.
"""
num_timesteps = len(all_timesteps_activations)
assert all([type(out) in (torch.Tensor, np.ndarray) for out in all_timesteps_activations]), \
"This function only takes all the activations for all the time steps of a single sample."
fig = plt.figure()
last_activations = all_timesteps_activations[0]
grid = AxesGrid(
fig,
111,
nrows_ncols=(1, num_timesteps - 1),
axes_pad=0.05,
share_all=True,
label_mode="L",
cbar_location="right",
cbar_mode="single",
)
for t, (axis, current_activations) in enumerate(
zip(grid, all_timesteps_activations[1:])):
activation_gradients = current_activations - last_activations
vmin, vmax = -2, 2
colormap = 'coolwarm'
if absolute:
vmin = 0
colormap = "Reds"
heatmap = axis.imshow(
activation_gradients.reshape(*neuron_heatmap_size),
cmap=colormap,
vmin=vmin,
vmax=vmax)
axis.set_xlabel("t={} -> t={}".format(t, t + 1))
axis.set_xticks([])
axis.set_yticks([])
last_activations = current_activations
grid.cbar_axes[0].colorbar(heatmap)
if show_title:
fig.suptitle("Activation value gradients over {} time steps".format(
num_timesteps))
if save is None:
plt.show()
else:
plt.savefig(save, bbox_inches="tight")
plt.close()
def accumulate_patches_into_heatmaps(self,
all_test_output,
outpath_prefix=''):
outpath = "plots/%s_%s.png" % (
outpath_prefix, path.splitext(path.basename(
self.test_imagepath))[0])
# http://matplotlib.org/examples/axes_grid/demo_axes_grid.html
fig = plt.figure()
grid = AxesGrid(
fig,
143, # similar to subplot(143)
nrows_ncols=(1, 1))
orig_img = imread(self.test_imagepath + '.png')
grid[0].imshow(orig_img)
grid = AxesGrid(
fig,
144, # similar to subplot(144)
nrows_ncols=(2, 2),
axes_pad=0.15,
label_mode="1",
share_all=True,
cbar_location="right",
cbar_mode="each",
cbar_size="7%",
cbar_pad="2%",
)
for klass in xrange(all_test_output.shape[1]):
accumulator = numpy.zeros(self.ds.image_shape[:2])
normalizer = numpy.zeros(self.ds.image_shape[:2])
for n in xrange(self.num_patch_centers):
i_start, i_end, j_start, j_end = self.nth_patch(n)
accumulator[i_start:i_end,
j_start:j_end] += all_test_output[n, klass]
normalizer[i_start:i_end, j_start:j_end] += 1
normalized_img = accumulator / normalizer
im = grid[klass].imshow(normalized_img,
interpolation="nearest",
vmin=0,
vmax=1)
grid.cbar_axes[klass].colorbar(im)
grid.axes_llc.set_xticks([])
grid.axes_llc.set_yticks([])
print("Saving figure as: %s" % outpath)
plt.savefig(outpath, dpi=600, bbox_inches='tight')
def plot_pathological_imgs():
fig = plt.figure()
grid = AxesGrid(fig, 111, nrows_ncols = (1, 4))
names = ['23050_right.png', '2468_left.png', '15450_left.png', '406_left.png']
imgs = [imread(n) for n in names]
[grid[i].imshow(imgs[i]) for i in range(len(imgs))]
plt.axis('off')
plt.savefig('out.png', dpi=300)
def main(args) :
import os.path
if args.trackTitles is None :
args.trackTitles = [os.path.dirname(filename) for
filename in args.inputDataFiles]
if len(args.inputDataFiles) == 0 :
print "WARNING: No corner control files given!"
if len(args.trackTitles) != len(args.inputDataFiles) :
raise ValueError("The number of TITLEs does not match the number"
" of INPUTFILEs.")
if args.statName is not None and args.statLonLat is None :
statData = ByName(args.statName)[0]
args.statLonLat = (statData['LON'], statData['LAT'])
if args.layout is None :
args.layout = (1, len(args.inputDataFiles))
if args.figsize is None :
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
if args.simTagFiles is None :
args.simTagFiles = []
polyfiles = args.polys
cornerVolumes = [ReadCorners(inFileName,
os.path.dirname(inFileName))['volume_data']
for inFileName in args.inputDataFiles]
polyData = [_load_verts(f, list(vol['stormCells'] for vol in vols)) for
f, vols in zip(polyfiles, cornerVolumes)]
multiTags = [ReadSimTagFile(fname) for fname in args.simTagFiles]
if len(multiTags) == 0 :
multiTags = [None]
if len(multiTags) < len(cornerVolumes) :
# Rudimentary broadcasting
tagMult = max(int(len(cornerVolumes) // len(multiTags)), 1)
multiTags = multiTags * tagMult
if args.statLonLat is not None :
for vols in cornerVolumes :
for vol in vols :
CoordinateTransform(vol['stormCells'],
args.statLonLat[0],
args.statLonLat[1])
for verts in polyData:
CoordinateTransform(verts,
args.statLonLat[0],
args.statLonLat[1])
showMap = (args.statLonLat is not None and args.displayMap)
showRadar = (args.statLonLat is not None and args.radarFile is not None)
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout,
share_all=True, axes_pad=0.32)
theAnim, radAnim = MakeCornerPlots(theFig, grid, cornerVolumes,
args.trackTitles, showMap, showRadar,
tail=args.tail,
startFrame=args.startFrame,
endFrame=args.endFrame,
radarFiles=args.radarFile,
fade=args.fade,
multiTags=multiTags,
tag_filters=args.filters)
polyAnims = []
for ax, verts in zip(grid, polyData):
from matplotlib.animation import ArtistAnimation
polyAnim = ArtistAnimation(theFig,
_to_polygons(polys[startFrame:endFrame + 1], ax),
event_source=theTimer)
polyAnims.append(polyAnim)
if args.xlims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None :
if radAnim is not None :
radAnim = [radAnim]
theAnim.save(args.saveImgFile, extra_anim=radAnim + polyAnims)
if args.doShow :
plt.show()
def main(args) :
import os.path # for os.path
import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import AxesGrid
inputDataFiles = []
titles = []
simTagFiles = []
if args.simName is not None :
dirName = os.path.join(args.directory, args.simName)
simParams = ParamUtils.ReadSimulationParams(os.path.join(dirName,
"simParams.conf"))
inputDataFiles.append(os.path.join(dirName, simParams['inputDataFile']))
titles.append(args.simName)
simTagFiles.append(os.path.join(dirName, simParams['simTagFile']))
# Add on any files specified at the command-line
inputDataFiles += args.inputDataFiles
titles += args.inputDataFiles
if args.simTagFiles is not None :
simTagFiles += args.simTagFiles
if len(inputDataFiles) == 0 :
print "WARNING: No inputDataFiles given or found!"
if len(titles) != len(inputDataFiles) :
raise ValueError("The number of TITLEs does not match the"
" number of INPUTFILEs.")
if len(simTagFiles) < len(inputDataFiles) :
# Not an error, just simply append None
simTagFiles.append([None] * (len(inputDataFiles) - len(simTagFiles)))
if args.statName is not None and args.statLonLat is None :
statData = ByName(args.statName)[0]
args.statLonLat = (statData['LON'], statData['LAT'])
if args.layout is None :
args.layout = (1, len(inputDataFiles))
if args.figsize is None :
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
cornerVolumes = [ReadCorners(inFileName,
os.path.dirname(inFileName))['volume_data']
for inFileName in inputDataFiles]
multiTags = [(ReadSimTagFile(fname) if fname is not None else None) for
fname in simTagFiles]
for vols, simTags in zip(cornerVolumes, multiTags) :
keeperIDs = process_tag_filters(simTags, args.filters)
if keeperIDs is None :
continue
for vol in vols :
vol['stormCells'] = FilterTrack(vol['stormCells'],
cornerIDs=keeperIDs)
if args.statLonLat is not None :
for vols in cornerVolumes :
for vol in vols :
CoordinateTransform(vol['stormCells'],
args.statLonLat[0],
args.statLonLat[1])
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout,
share_all=True, axes_pad=0.32)
# A list to hold the CircleCollection arrays, it will have length
# of max(tLims) - min(tLims) + 1
allCorners = None
if args.trackFile is not None :
(tracks, falarms) = FilterMHTTracks(*ReadTracks(args.trackFile))
if args.statLonLat is not None :
CoordinateTransform(tracks + falarms,
args.statLonLat[0],
args.statLonLat[1])
(xLims, yLims, frameLims) = DomainFromTracks(tracks + falarms)
else :
volumes = []
for aVol in cornerVolumes :
volumes.extend(aVol)
(xLims, yLims, tLims, frameLims) = DomainFromVolumes(volumes)
showMap = (args.statLonLat is not None and args.displayMap)
if showMap :
bmap = Basemap(projection='cyl', resolution='l',
suppress_ticks=False,
llcrnrlat=yLims[0], llcrnrlon=xLims[0],
urcrnrlat=yLims[1], urcrnrlon=xLims[1])
startFrame = args.startFrame
endFrame = args.endFrame
tail = args.tail
if startFrame is None :
startFrame = frameLims[0]
if endFrame is None :
endFrame = frameLims[1]
if tail is None :
tail = 0
# A common event_source for synchronizing all the animations
theTimer = None
# Make the corners big
big = False
if args.radarFile is not None and args.statLonLat is not None :
if endFrame - frameLims[0] >= len(args.radarFile) :
# Not enough radar files, so truncate the tracks.
endFrame = (len(args.radarFile) + frameLims[0]) - 1
files = args.radarFile[startFrame - frameLims[0]:(endFrame + 1) -
frameLims[0]]
radAnim = RadarAnim(theFig, files)
theTimer = radAnim.event_source
for ax in grid :
radAnim.add_axes(ax, alpha=0.6, zorder=0)
# Radar images make it difficult to see corners, so make 'em big
big = True
else :
radAnim = None
theAnim = CornerAnimation(theFig, endFrame - startFrame + 1,
tail=tail, interval=250, blit=False,
event_source=theTimer, fade=args.fade)
for (index, volData) in enumerate(cornerVolumes) :
curAxis = grid[index]
if showMap :
PlotMapLayers(bmap, mapLayers, curAxis, zorder=0.1)
volFrames = [frameVol['frameNum'] for frameVol in volData]
startIdx = volFrames.index(startFrame)
endIdx = volFrames.index(endFrame)
volTimes = [frameVol['volTime'] for frameVol in volData]
startT = volTimes[startIdx]
endT = volTimes[endIdx]
corners = PlotCorners(volData, (startT, endT), axis=curAxis,
big=big)
#curAxis.set_aspect("equal", 'datalim')
#curAxis.set_aspect("equal")
curAxis.set_title(titles[index])
if not showMap :
curAxis.set_xlabel("X")
curAxis.set_ylabel("Y")
else :
curAxis.set_xlabel("Longitude")
curAxis.set_ylabel("Latitude")
theAnim.AddCornerVolume(corners)
if args.xlims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None :
if radAnim is not None :
radAnim = [radAnim]
theAnim.save(args.saveImgFile, extra_anim=radAnim)
if args.doShow :
plt.show()
def main(args) :
import os.path
import glob # for globbing
import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import AxesGrid
if args.bw_mode :
BW_mode() # from TrackPlot module
# FIXME: Currently, the code allows for trackFiles to be listed as well
# as providing a simulation (which trackfiles are automatically
# grabbed). Both situations can not be handled right now, though.
trackFiles = []
trackTitles = []
if args.simName is not None :
dirName = os.path.join(args.directory, args.simName)
simParams = ParamUtils.ReadSimulationParams(os.path.join(dirName,
"simParams.conf"))
if args.trackRuns is not None :
simParams['trackers'] = ExpandTrackRuns(simParams['trackers'],
args.trackRuns)
trackFiles = [os.path.join(dirName, simParams['result_file'] +
'_' + aTracker)
for aTracker in simParams['trackers']]
if args.trackTitles is None :
trackTitles = simParams['trackers']
else :
trackTitles = args.trackTitles
if args.truthTrackFile is None :
args.truthTrackFile = os.path.join(dirName,
simParams['noisyTrackFile'])
if args.simTagFile is None :
args.simTagFile = os.path.join(dirName, simParams['simTagFile'])
trackFiles += args.trackFiles
if args.trackTitles is None :
trackTitles += args.trackFiles
else :
trackTitles += args.trackTitles
if len(trackFiles) == 0 : print "WARNING: No trackFiles given or found!"
if len(trackTitles) != len(trackFiles) :
raise ValueError("The number of TITLEs do not match the"
" number of TRACKFILEs.")
if args.statName is not None and args.statLonLat is None :
statData = ByName(args.statName)[0]
args.statLonLat = (statData['LON'], statData['LAT'])
if args.layout is None :
args.layout = (1, len(trackFiles))
if args.figsize is None :
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
trackerData = [FilterMHTTracks(*ReadTracks(trackFile)) for
trackFile in trackFiles]
keeperIDs = None
if args.simTagFile is not None :
simTags = ParamUtils.ReadSimTagFile(args.simTagFile)
keeperIDs = ParamUtils.process_tag_filters(simTags, args.filters)
if args.statLonLat is not None :
for aTracker in trackerData :
CoordinateTransform(aTracker[0] + aTracker[1],
args.statLonLat[0],
args.statLonLat[1])
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout, aspect=False,
share_all=True, axes_pad=0.35)
if args.truthTrackFile is not None :
(true_tracks,
true_falarms) = FilterMHTTracks(*ReadTracks(args.truthTrackFile))
if args.statLonLat is not None :
CoordinateTransform(true_tracks + true_falarms,
args.statLonLat[0],
args.statLonLat[1])
true_AssocSegs = CreateSegments(true_tracks)
true_FAlarmSegs = CreateSegments(true_falarms)
if keeperIDs is not None :
true_AssocSegs = FilterSegments(keeperIDs, true_AssocSegs)
true_FAlarmSegs = FilterSegments(keeperIDs, true_FAlarmSegs)
(xLims, yLims, frameLims) = DomainFromTracks(true_tracks + true_falarms)
else :
true_AssocSegs = None
true_FAlarmSegs = None
stackedTracks = []
for aTracker in trackerData :
stackedTracks += aTracker[0] + aTracker[1]
(xLims, yLims, frameLims) = DomainFromTracks(stackedTracks)
endFrame = args.endFrame
tail = args.tail
if endFrame is None :
endFrame = frameLims[1]
if tail is None :
tail = endFrame - frameLims[0]
startFrame = endFrame - tail
showMap = (args.statLonLat is not None and args.displayMap)
if args.radarFile is not None and args.statLonLat is not None :
if len(args.radarFile) > 1 and args.endFrame is not None :
args.radarFile = args.radarFile[args.endFrame]
else :
args.radarFile = args.radarFile[-1]
raddata = LoadRastRadar(args.radarFile)
else :
raddata = None
if showMap :
bmap = Basemap(projection='cyl', resolution='i',
suppress_ticks=False,
llcrnrlat=yLims[0], llcrnrlon=xLims[0],
urcrnrlat=yLims[1], urcrnrlon=xLims[1])
for index, (tracks, falarms) in enumerate(trackerData) :
curAxis = grid[index]
if raddata is not None :
MakeReflectPPI(raddata['vals'][0], raddata['lats'], raddata['lons'],
meth='pcmesh', ax=curAxis, colorbar=False,
axis_labels=False, zorder=0, alpha=0.6)
if showMap :
PlotMapLayers(bmap, mapLayers, curAxis)
if true_AssocSegs is not None and true_FAlarmSegs is not None :
trackAssocSegs = CreateSegments(tracks)
trackFAlarmSegs = CreateSegments(falarms)
if keeperIDs is not None :
trackAssocSegs = FilterSegments(keeperIDs, trackAssocSegs)
trackFAlarmSegs = FilterSegments(keeperIDs, trackFAlarmSegs)
truthtable = CompareSegments(true_AssocSegs, true_FAlarmSegs,
trackAssocSegs, trackFAlarmSegs)
PlotSegments(truthtable, (startFrame, endFrame), axis=curAxis,
fade=args.fade)
else :
if keeperIDs is not None :
filtFunc = lambda trk: FilterTrack(trk, cornerIDs=keeperIDs)
tracks = map(filtFunc, tracks)
falarms = map(filtFunc, falarms)
CleanupTracks(tracks, falarms)
PlotPlainTracks(tracks, falarms,
startFrame, endFrame, axis=curAxis,
fade=args.fade)
#curAxis.set_xlim(xLims)
#curAxis.set_ylim(yLims)
#curAxis.set_aspect("equal", 'datalim')
#curAxis.set_aspect("equal")
curAxis.set_title(trackTitles[index])
if not showMap :
curAxis.set_xlabel("X")
curAxis.set_ylabel("Y")
else :
curAxis.set_xlabel("Longitude")
curAxis.set_ylabel("Latitude")
if args.xlims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None :
theFig.savefig(args.saveImgFile, bbox_inches='tight')
if args.doShow :
plt.show()
def main(args) :
if args.bw_mode :
BW_mode()
if len(args.trackFiles) == 0 :
print "WARNING: No trackFiles given!"
if len(args.truthTrackFile) == 0 :
print "WARNING: No truth trackFiles given!"
if args.trackTitles is None :
args.trackTitles = args.trackFiles
else :
if len(args.trackTitles) != len(args.trackFiles) :
raise ValueError("The number of TITLEs does not match the number"
" of TRACKFILEs")
if args.statName is not None and args.statLonLat is None :
statData = ByName(args.statName)[0]
args.statLonLat = (statData['LON'], statData['LAT'])
if args.layout is None :
args.layout = (1, max(len(args.trackFiles),
len(args.truthTrackFile)))
if args.figsize is None :
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
if args.simTagFiles is None :
args.simTagFiles = []
trackerData = [FilterMHTTracks(*ReadTracks(trackFile)) for
trackFile in args.trackFiles]
truthData = [FilterMHTTracks(*ReadTracks(trackFile)) for
trackFile in args.truthTrackFile]
multiTags = [ReadSimTagFile(fname) for fname in args.simTagFiles]
if len(multiTags) == 0 :
multiTags = [None]
if args.statLonLat is not None :
for aTracker in trackerData + truthData :
CoordinateTransform(aTracker[0] + aTracker[1],
args.statLonLat[0],
args.statLonLat[1])
if len(trackerData) != len(truthData) :
# Basic broadcasting needed!
if len(truthData) > len(trackerData) :
# Need to extend track data to match with the number of truth sets
if len(truthData) % len(trackerData) != 0 :
raise ValueError("Can't extend TRACKFILE list to match with"
" the TRUTHFILE list!")
else :
# Need to extend truth sets to match with the number of track data
if len(trackerData) % len(truthData) != 0 :
raise ValueError("Can't extend TRUTHFILE list to match with"
" the TRACKFILE list!")
trkMult = max(int(len(truthData) // len(trackerData)), 1)
trthMult = max(int(len(trackerData) // len(truthData)), 1)
trackerData = trackerData * trkMult
truthData = truthData * trthMult
tagMult = max(int(len(truthData) // len(multiTags)), 1)
multiTags = multiTags * tagMult
args.trackTitles = args.trackTitles * trkMult
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout, aspect=False,
share_all=True, axes_pad=0.45)
showMap = (args.statLonLat is not None and args.displayMap)
if args.radarFile is not None and args.statLonLat is not None :
if len(args.radarFile) > 1 and args.endFrame is not None :
args.radarFile = args.radarFile[args.endFrame]
else :
args.radarFile = args.radarFile[-1]
data = LoadRastRadar(args.radarFile)
for ax in grid :
MakeReflectPPI(data['vals'][0], data['lats'], data['lons'],
meth='pcmesh', ax=ax, colorbar=False,
axis_labels=False, zorder=0, alpha=0.6)
MakeComparePlots(grid, trackerData, truthData, args.trackTitles, showMap,
endFrame=args.endFrame, tail=args.tail, fade=args.fade,
multiTags=multiTags, tag_filters=args.filters)
if args.xlims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None :
theFig.savefig(args.saveImgFile)
if args.doShow :
plt.show()
def main(args) :
import os.path # for os.path.join()
import glob # for globbing
if args.bw_mode :
BW_mode() # from TrackPlot module
# FIXME: Currently, the code allows for trackFiles to be listed as well
# as providing a simulation (which trackfiles are automatically
# grabbed). Both situations can not be handled right now, though.
trackFiles = []
trackTitles = []
polyfiles = args.polys
if args.statName is not None and args.statLonLat is None :
statData = ByName(args.statName)[0]
args.statLonLat = (statData['LON'], statData['LAT'])
if args.simName is not None :
dirName = os.path.join(args.directory, args.simName)
simParams = ParamUtils.ReadSimulationParams(os.path.join(dirName,
"simParams.conf"))
if args.trackRuns is not None :
simParams['trackers'] = ExpandTrackRuns(simParams['trackers'],
args.trackRuns)
trackFiles = [os.path.join(dirName, simParams['result_file'] +
'_' + aTracker)
for aTracker in simParams['trackers']]
trackTitles = simParams['trackers']
if args.truthTrackFile is None :
args.truthTrackFile = os.path.join(dirName,
simParams['noisyTrackFile'])
if args.simTagFile is None :
args.simTagFile = os.path.join(dirName,
simParams['simTagFile'])
trackFiles += args.trackFiles
trackTitles += args.trackFiles
if args.trackTitles is not None :
trackTitles = args.trackTitles
if len(trackFiles) == 0 : print "WARNING: No trackFiles given or found!"
if args.layout is None :
args.layout = (1, len(trackFiles))
if args.figsize is None :
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
if len(trackFiles) < len(polyfiles):
raise ValueError("Can not have more polygon files than trackfiles!")
trackerData = [FilterMHTTracks(*ReadTracks(trackFile)) for
trackFile in trackFiles]
polyData = [_load_verts(f, tracks + falarms) for f, (tracks, falarms) in
zip(polyfiles, trackerData)]
keeperIDs = None
if args.simTagFile is not None :
simTags = ParamUtils.ReadSimTagFile(args.simTagFile)
keeperIDs = ParamUtils.process_tag_filters(simTags, args.filters)
if args.statLonLat is not None :
for aTracker in trackerData :
CoordinateTransform(aTracker[0] + aTracker[1],
args.statLonLat[0],
args.statLonLat[1])
for polys in polyData:
CoordinateTrans_lists(polys,
args.statLonLat[0], args.statLonLat[1])
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout,# aspect=False,
share_all=True, axes_pad=0.45)
if args.truthTrackFile is not None :
(true_tracks,
true_falarms) = FilterMHTTracks(*ReadTracks(args.truthTrackFile))
if args.statLonLat is not None :
CoordinateTransform(true_tracks + true_falarms,
args.statLonLat[0],
args.statLonLat[1])
true_AssocSegs = CreateSegments(true_tracks)
true_FAlarmSegs = CreateSegments(true_falarms)
if keeperIDs is not None :
true_AssocSegs = FilterSegments(keeperIDs, true_AssocSegs)
true_FAlarmSegs = FilterSegments(keeperIDs, true_FAlarmSegs)
(xLims, yLims, frameLims) = DomainFromTracks(true_tracks + true_falarms)
else :
true_AssocSegs = None
true_FAlarmSegs = None
stackedTracks = []
for aTracker in trackerData :
stackedTracks += aTracker[0] + aTracker[1]
(xLims, yLims, frameLims) = DomainFromTracks(stackedTracks)
startFrame = args.startFrame
endFrame = args.endFrame
tail = args.tail
if startFrame is None :
startFrame = 0
if endFrame is None :
endFrame = frameLims[1]
if tail is None :
tail = endFrame - startFrame
# A common timer for all animations for syncing purposes.
theTimer = None
if args.radarFile is not None and args.statLonLat is not None :
if endFrame >= len(args.radarFile) :
# Not enough radar files, so truncate the tracks.
endFrame = len(args.radarFile) - 1
files = args.radarFile[startFrame:(endFrame + 1)]
radAnim = RadarAnim(theFig, files)
theTimer = radAnim.event_source
for ax in grid :
radAnim.add_axes(ax, alpha=0.6, zorder=0)
else :
radAnim = None
showMap = (args.statLonLat is not None and args.displayMap)
if showMap :
bmap = Basemap(projection='cyl', resolution='i',
suppress_ticks=False,
llcrnrlat=yLims[0], llcrnrlon=xLims[0],
urcrnrlat=yLims[1], urcrnrlon=xLims[1])
animator = SegAnimator(theFig, startFrame, endFrame, tail,
event_source=theTimer, fade=args.fade)
for index, (tracks, falarms) in enumerate(trackerData):
curAxis = grid[index]
if showMap :
PlotMapLayers(bmap, mapLayers, curAxis, zorder=0.1)
if true_AssocSegs is not None and true_FAlarmSegs is not None :
trackAssocSegs = CreateSegments(tracks)
trackFAlarmSegs = CreateSegments(falarms)
if keeperIDs is not None :
trackAssocSegs = FilterSegments(keeperIDs, trackAssocSegs)
trackFAlarmSegs = FilterSegments(keeperIDs, trackFAlarmSegs)
truthtable = CompareSegments(true_AssocSegs, true_FAlarmSegs,
trackAssocSegs, trackFAlarmSegs)
l, d = Animate_Segments(truthtable, (startFrame, endFrame),
axis=curAxis)
else :
if keeperIDs is not None :
filtFunc = lambda trk : FilterTrack(trk, cornerIDs=keeperIDs)
tracks = map(filtFunc, tracks)
falarms = map(filtFunc, falarms)
CleanupTracks(tracks, falarms)
l, d = Animate_PlainTracks(tracks, falarms,
(startFrame, endFrame), axis=curAxis)
animator._lines.extend(l)
animator._lineData.extend(d)
#curAxis.set_aspect("equal", 'datalim')
#curAxis.set_aspect("equal")
curAxis.set_title(trackTitles[index])
if not showMap :
curAxis.set_xlabel("X")
curAxis.set_ylabel("Y")
else :
curAxis.set_xlabel("Longitude (degrees)")
curAxis.set_ylabel("Latitude (degrees)")
polyAnims = []
for ax, verts in zip(grid, polyData):
from matplotlib.animation import ArtistAnimation
polyAnim = ArtistAnimation(theFig,
_to_polygons(polys[startFrame:endFrame + 1], ax),
event_source=theTimer)
polyAnims.append(polyAnim)
if args.xlims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0 :
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None :
if radAnim is not None :
radAnim = [radAnim]
else:
radAnim = []
animator.save(args.saveImgFile, extra_anim=radAnim + polyAnims)
if args.doShow :
plt.show()
def main(args):
if args.bw_mode:
BW_mode()
if len(args.trackFiles) == 0:
print "WARNING: No trackFiles given!"
if args.trackTitles is None:
args.trackTitles = args.trackFiles
else:
if len(args.trackTitles) != len(args.trackFiles):
raise ValueError("The number of TITLEs do not match the" " number of TRACKFILEs.")
if args.statName is not None and args.statLonLat is None:
statData = ByName(args.statName)[0]
args.statLonLat = (statData["LON"], statData["LAT"])
if args.layout is None:
args.layout = (1, len(args.trackFiles))
if args.figsize is None:
args.figsize = plt.figaspect(float(args.layout[0]) / args.layout[1])
trackerData = [FilterMHTTracks(*ReadTracks(trackFile)) for trackFile in args.trackFiles]
if args.statLonLat is not None:
for aTracker in trackerData:
CoordinateTransform(aTracker[0] + aTracker[1], args.statLonLat[0], args.statLonLat[1])
if args.simTagFiles is None:
args.simTagFiles = [None]
multiTags = [ReadSimTagFile(fname) if fname is not None else None for fname in args.simTagFiles]
if len(trackerData) > len(multiTags):
# Very rudimentary broadcasting of multiTags to match trackerData
tagMult = max(int(len(trackerData) // len(multiTags)), 1)
multiTags = multiTags * tagMult
theFig = plt.figure(figsize=args.figsize)
grid = AxesGrid(theFig, 111, nrows_ncols=args.layout, aspect=False, share_all=True, axes_pad=0.45)
showMap = args.statLonLat is not None and args.displayMap
# Can only do this if all other data being displayed will be in
# lon/lat coordinates
if args.radarFile is not None and args.statLonLat is not None:
if len(args.radarFile) > 1 and args.endFrame is not None:
args.radarFile = args.radarFile[args.endFrame]
else:
args.radarFile = args.radarFile[-1]
data = LoadRastRadar(args.radarFile)
for ax in grid:
MakeReflectPPI(
data["vals"][0],
data["lats"],
data["lons"],
meth="pcmesh",
ax=ax,
colorbar=False,
axis_labels=False,
zorder=0,
alpha=0.6,
)
MakeTrackPlots(
grid,
trackerData,
args.trackTitles,
showMap,
endFrame=args.endFrame,
tail=args.tail,
fade=args.fade,
multiTags=multiTags,
tag_filters=args.filters,
)
if args.xlims is not None and np.prod(grid.get_geometry()) > 0:
grid[0].set_xlim(args.xlims)
if args.ylims is not None and np.prod(grid.get_geometry()) > 0:
grid[0].set_ylim(args.ylims)
if args.saveImgFile is not None:
theFig.savefig(args.saveImgFile)
if args.doShow:
plt.show()
