def output_fig(self,figure_filename,xlabel='Temp [degC]',display=True):
months=list(self.profile.keys())
number_of_loops=len(months)
spec = strategies.SquareStrategy().get_grid(number_of_loops)
fig = plt.gcf()
fig.set_dpi(100)
fig.constrained_layout=True
fig.set_figheight(11.69)
fig.set_figwidth(8.27)
xmin=np.inf
xmax=-np.inf
for month in months:
xmin=min(xmin,np.nanmin(self.profile[month]))
xmax=max(xmax,np.nanmax(self.profile[month]))
for j,sub in enumerate(spec):
ax = plt.subplot(sub)
ax.plot(self.profile[months[j]],self.x*-1)
ax.set_xlabel(xlabel)
ax.set_ylabel('Water depth (m)')
ax.set_title(months[j])
ax.set_xlim(xmin,xmax)
fig.align_labels()
if display:
plt.show(block=~display)
plt.savefig(figure_filename)
def plot(subunits):
"""Plots actuators
Arguments:
subunits {list} -- List of subunit ORIENTATION VECTORS
"""
num_plots = len(subunits)
specs = strategies.SquareStrategy('center').get_grid(num_plots)
fig = plt.figure(1)
for vec, sub in zip(subunits, specs):
ax = fig.add_subplot(sub)
l = Limb()
l.build(vec)
ax.plot([0, 0], [-2, 2], color='black')
ax.plot(l.XY[0], l.XY[1], color='red')
ax.set_aspect('equal', adjustable='datalim')
ax.margins(1.5, 1.5)
ax.autoscale(False)
overlay_images(ax, l)
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
plt.show()
def show_image_grid(images):
import matplotlib.pyplot as plt
from grid_strategy import strategies
batch_size = images.shape[0]
specs = strategies.SquareStrategy("center").get_grid(batch_size)
for b, subplot in enumerate(specs):
plt.subplot(subplot)
plt.xticks([])
plt.yticks([])
ax = plt.gca()
if b == 0:
ax.set_title(r"$I_t$")
ax.imshow(images[b, 0, :, :, :3])
elif b == 1:
ax.set_title(r"$I_{t+1}$")
ax.imshow(images[b - 1, 1, :, :, :3])
else:
ax.imshow(images[b - 1, 1, :, :, :3])
plt.show()
def square_strategy():
return strategies.SquareStrategy()
def plot_limb(limbs, objt=None, curve=None):
for i in range(len(limbs[0])):
if isinstance(limbs[0][i], list):
vec_ind = i
limbs = np.array(limbs)
if len(limbs.shape) > 1:
num_plots = len(limbs)
limbs = limbs[:, vec_ind]
else:
num_plots = len(limbs.shape)
limbs = [limbs]
specs = strategies.SquareStrategy('center').get_grid(num_plots)
fig = plt.figure(1, constrained_layout=False)
fig.canvas.set_window_title('Actuator')
for vec, sub in zip(limbs, specs):
ax = fig.add_subplot(sub)
limb = Limb()
limb.build(vec)
end = np.array((limb.XY[0, -num_points:], limb.XY[1, -num_points:]))
segs = len(limb.orient)
coordinates = np.copy(limb.XY)
# ax.set_title("Soft actuator\n" + "Number of segments: {}".format(segs))
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
ax.plot([0, 0], [-2, 2], color='black')
ax.xaxis.set_major_locator(MultipleLocator(1))
normal = np.zeros((len(vec) + 1))
''' PLOT UNACTUATED LINE '''
# ax.plot(normal, color='grey',
# label="Initial pressure (P=P" + r'$_i$' + ")")
"""------ACTUATED-------"""
ax.plot(coordinates[0, :],
coordinates[1, :],
color='red',
label="Final pressure (P=P" + r'$_f$' + ")",
alpha=0.6)
if isinstance(objt, np.ndarray):
ax.plot(objt[0], objt[1], color='black')
elif isinstance(objt, Polygon):
ax.plot(objt.exterior.xy[0], objt.exterior.xy[1], color='black')
ax.plot(curve[0], curve[1], color='blue', alpha=0.6)
''' VISUALISATION OF CURVE FIT '''
if zdist(end, curve) > zdist(end, curve[::-1]):
ax.plot([end[0], curve[0][::-1]], [end[1], curve[1][::-1]],
linewidth=2,
color='green')
else:
ax.plot([end[0], curve[0]], [end[1], curve[1]],
linewidth=2,
color='green')
ax.margins(0.5, 0.5)
ax.set_aspect('equal', adjustable='datalim')
ax.autoscale(False)
overlay_images(ax, limb)
plt.tight_layout()
plt.show()
def plot_limb(limbs):
def on_click(event):
ax = event.inaxes
if ax is None:
return
if event.button != 1:
return
if zoomed_axes[0] is None:
zoomed_axes[0] = (ax, ax.get_position())
ax.set_position([0.1, 0.1, 0.8, 0.8])
ax.legend(loc='best')
ax.get_xaxis().set_visible(True)
ax.get_yaxis().set_visible(True)
ax.xaxis.set_major_locator(MultipleLocator(5))
ax.grid(linestyle=':')
ax.set_ylim(0)
ax.margins(x=0, y=-0.25)
for axis in event.canvas.figure.axes:
if axis is not ax:
axis.set_visible(False)
else:
zoomed_axes[0][0].set_position(zoomed_axes[0][1])
zoomed_axes[0] = None
ax.get_legend().remove()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
for axis in event.canvas.figure.axes:
axis.set_visible(True)
event.canvas.draw()
zoomed_axes = [None]
for i in range(len(limbs[0])):
if isinstance(limbs[0][i], list):
vec_ind = i
limbs = np.array(limbs)
if len(limbs.shape) > 1:
num_plots = len(limbs)
limbs = limbs[:, vec_ind]
else:
num_plots = len(limbs.shape)
limbs = [limbs]
# fig = vp.Fig(size=(600,500), show=False)
# limb = Limb()
# limb.build(limbs[0])
# points = np.copy(limb.XY)
# line = fig[0, 0].plot((points[0, :], points[1, :]), color='red')
# fig.show(run=True)
specs = strategies.SquareStrategy('center').get_grid(num_plots)
fig = plt.figure(1, constrained_layout=False)
# fig.canvas.set_window_title('Top ' + str(num_plots))
for vec, sub in zip(limbs, specs):
ax = fig.add_subplot(sub)
limb = Limb()
limb.build(vec)
segs = len(limb.orient)
points = np.copy(limb.XY)
rots = limb.curvature
# ax.set_title("Soft actuator\n" + "Number of segments: {}".format(segs))
# ax.get_xaxis().set_visible(False)
# ax.get_yaxis().set_visible(False)
ax.plot([0, 0], [-2, 2], color='black')
# ax.xaxis.set_major_locator(MultipleLocator(1))
if settings.get('Rainbow'):
colors = cm.rainbow(np.linspace(0, 1, segs))
"""------NORMAL-------"""
for i in range(0, segs - 1):
ax.plot([i, i + 2], [0, 0], color=colors[i])
"""------ACTUATED-------"""
for i in range(0, segs - 1):
ax.plot(points[0, i:i + 2],
points[1, i:i + 2],
color=colors[i])
else:
"""------NORMAL-------"""
# normal = np.zeros((segs+1))
# ax.plot(normal, color='grey',
# label="Initial pressure (P=P" + r'$_i$' + ")")
"""------ACTUATED-------"""
ax.plot(points[0, :],
points[1, :],
color='red',
label="Reduced-order model")
if any(value == True
for value in parameters.get('Curve fitting').values()):
m = points[0][-1] / (2 * pi)
if parameters.get('Curve fitting').get('Sin'):
curve = 20 * np.sin(points[0] / m)
elif parameters.get('Curve fitting').get('Cos'):
curve = 25 * np.cos(points[0] / m) - 25
elif parameters.get('Curve fitting').get('Custom'):
curve = []
func = parameters.get('Curve fitting').get('Custom func')
for ea in points[0] / m:
x = ea
curve.append(eval(func))
ax.plot(points[0],
curve,
color='black',
alpha=0.85,
linestyle='--',
label='Desired profile')
# ax.margins(0.5, 0.5)
# ax.set_xlim(0)
# ax.margins(x=0, y=-0.25)
if settings.get('Plot boundaries'):
to_tuple = [(x, y) for x, y in zip(limb.XY[0], limb.XY[1])]
line_check = LineString(to_tuple)
line_top = line_check.parallel_offset(1.9, side='left')
line_bottom = line_check.parallel_offset(1.9, side='right')
ax.plot(line_top.xy[0], line_top.xy[1])
ax.plot(line_bottom.xy[0], line_bottom.xy[1])
if settings.get('Overlay images'):
def imshow_affine(ax, z, *args, **kwargs):
im = ax.imshow(z, *args, **kwargs)
_, x2, y1, _ = im.get_extent()
im._image_skew_coordinate = (x2, y1)
return im
# width = 2.57
# height = 3.9
width = 22.5
height = 35
image_directory = os.path.dirname(
os.path.realpath(__file__)) + '\\box.png'
# img = plt.imread(image_directory, format='png')
# img = Image.open(image_directory)
img = mpimg.imread(image_directory)
cps = [[], []]
for i in range(segs):
cps[0].append((points[0][i] + points[0][i + 1]) / 2)
cps[1].append((points[1][i] + points[1][i + 1]) / 2)
cps = np.asarray(cps)
for i in range(cps.shape[1]):
img_show = imshow_affine(
ax,
img,
interpolation='none',
extent=[0, width, 0, height],
)
c_x, c_y = width / 2, (16 * cos(radians(11)))
if limb.orient[i] == "TOP":
rot_angle = 180 + degrees(rots[i + 1])
elif limb.orient[i] == "BOTTOM":
rot_angle = degrees(rots[i + 1])
else:
rot_angle = degrees(rots[i + 1])
transform_data = (transforms.Affine2D().rotate_deg_around(
c_x, c_y, rot_angle).translate(
(cps[0][i] - c_x), (cps[1][i] - c_y)) + ax.transData)
img_show.set_transform(transform_data)
diff = 20
x_min = min(points[0, :]) - diff
x_max = max(points[0, :]) + diff
y_min = min(points[1, :]) - diff
y_max = max(points[1, :]) + diff
# fig.canvas.mpl_connect('button_press_event', on_click)
# ax.set_aspect('equal', adjustable='datalim')
# ax.autoscale()
# plt.tight_layout()
plt.xlim(x_min, x_max)
plt.ylim(y_min, y_max)
plt.xlabel('X [mm]')
plt.ylabel('Y [mm]')
ax.xaxis.set_major_locator(MultipleLocator(20))
ax.grid(linestyle=':')
# ax.margins(0.5, 0.1)
plotDavid(ax)
figManager = plt.get_current_fig_manager()
figManager.window.state('zoomed')
lgd = ax.legend(loc='lower center', bbox_to_anchor=(0.5, 1))
# plt.show()
plt.savefig('C:\\Users\\zjmon\\Documents\\Meesters\\Thesis\\figs\\Cos.pgf')
def do_perc_of_occurence(time, mag, drr, mag_interval, xlabel, time_blocking,
dir_int, fileout, show):
## Input
gd_value = ~np.isnan(mag)
time = time[gd_value]
mag = mag[gd_value]
if drr is not None:
drr = drr[gd_value]
else:
drr = np.ones((len(mag), ))
dir_int = [0, 360]
if isinstance(mag_interval, int) or isinstance(mag_interval, float):
s = np.arange(0, np.max(mag), mag_interval)
elif isinstance(mag_interval, list):
if len(mag_interval) < 2:
s = np.linspace(0, np.max(mag), 10)
else:
s = np.array(mag_interval)
else:
s = np.linspace(0, np.max(mag), 10)
month = time.month
number_of_loops, identifiers, month_identifier = get_number_of_loops(
time_blocking)
index = []
nj = []
for j in range(0, number_of_loops):
tmp = np.in1d(month, month_identifier[j])
if np.any(tmp):
#index .append(tmp)
nj.append(j)
number_of_real_loops = len(nj)
spec = strategies.SquareStrategy().get_grid(number_of_real_loops)
fig = plt.gcf()
fig.set_dpi(100)
fig.constrained_layout = True
fig.set_figheight(11.69)
fig.set_figwidth(8.27)
for i, sub in enumerate(spec):
ax = plt.subplot(sub)
#Pull out relevant indices for particular month/months
index = np.in1d(month, month_identifier[nj[i]])
big_length = len(index.nonzero()[0])
if big_length > 0:
SPD = mag[index]
DIR = drr[index]
jet = cm = plt.get_cmap('jet')
cNorm = colors.Normalize(vmin=0, vmax=len(dir_int) - 2)
scalarMap = cmx.ScalarMappable(norm=cNorm, cmap=jet)
for jj in range(0, len(dir_int) - 1):
if dir_int[jj + 1] <= dir_int[jj]:
D = np.logical_or(
np.mod(DIR, 360) > dir_int[jj],
np.mod(DIR, 360) <= dir_int[jj + 1])
else:
D = (np.mod(DIR, 360) > dir_int[jj]) & (np.mod(DIR, 360) <=
dir_int[jj + 1])
colorVal = scalarMap.to_rgba(jj)
perc = get_perc(s, SPD[D]) / big_length
plt.plot(s[:-1] + np.diff(s) / 2,
perc,
color=colorVal,
label=degToCompass([dir_int[jj], dir_int[jj + 1]]))
ax.set_title(identifiers[nj[i]])
if i == number_of_real_loops - 1 and len(dir_int) > 2:
# if number_of_loops>3 and number_of_loops<10:
# ax.legend(loc='best',bbox_to_anchor=(0.6, -0.4),ncol=len(dir_int)-1)#bbox_to_anchor=(0.8,-1.0, 0.5, 0.5))
# elif number_of_loops>10:
# ax.legend(loc='best',bbox_to_anchor=(0.6, -3.4),ncol=len(dir_int)-1)#bbox_to_anchor=(0.8,-1.0, 0.5, 0.5))
# else:
ax.legend(loc='best')
#if int(y)==0:
ax.set_ylabel('% Occurence')
#if int(x)==maxx :
ax.set_xlabel(xlabel)
fig.align_labels()
# if number_of_loops>10:
# plt.subplots_adjust(left=0.075,right=0.970,bottom=0.075,top=0.97,hspace=.5,wspace=0.415)
# ax.set_xlabel(xlabel)
# elif number_of_loops>2 and number_of_loops<10:
# plt.subplots_adjust(left=0.08,right=0.975,bottom=0.05,top=0.7,hspace=.5,wspace=0.3)
# ax.set_xlabel(xlabel)
# else:
# plt.subplots_adjust(bottom=0.05,top=.95,hspace=.5)
if show:
plt.show(block=~show)
plt.savefig(fileout)
def plot_limb(limbs):
def on_click(event):
ax = event.inaxes
if ax is None:
return
if event.button != 1:
return
if zoomed_axes[0] is None:
zoomed_axes[0] = (ax, ax.get_position())
ax.set_position([0.1, 0.1, 0.8, 0.8])
ax.legend(loc='best')
ax.get_xaxis().set_visible(True)
ax.get_yaxis().set_visible(True)
ax.xaxis.set_major_locator(MultipleLocator(5))
ax.grid(linestyle=':')
ax.set_ylim(0)
ax.margins(x=0, y=-0.25)
for axis in event.canvas.figure.axes:
if axis is not ax:
axis.set_visible(False)
else:
zoomed_axes[0][0].set_position(zoomed_axes[0][1])
zoomed_axes[0] = None
ax.get_legend().remove()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
for axis in event.canvas.figure.axes:
axis.set_visible(True)
event.canvas.draw()
zoomed_axes = [None]
for i in range(len(limbs[0])):
if isinstance(limbs[0][i], list):
vec_ind = i
limbs = np.array(limbs)
if len(limbs.shape) > 1:
num_plots = len(limbs)
limbs = limbs[:, vec_ind]
else:
num_plots = len(limbs.shape)
limbs = [limbs]
specs = strategies.SquareStrategy('center').get_grid(num_plots)
fig = plt.figure(1, constrained_layout=False)
fig.canvas.set_window_title('Top ' + str(num_plots))
for vec, sub in zip(limbs, specs):
ax = fig.add_subplot(sub)
limb = Limb()
limb.build(vec)
segs = len(limb.orient)
points = np.copy(limb.XY)
rots = limb.curvature
ax.plot([0, 0], [-2, 2], color='black')
if settings.get('Rainbow'):
colors = cm.rainbow(np.linspace(0, 1, segs))
"""------NORMAL-------"""
for i in range(0, segs - 1):
ax.plot([i, i + 2], [0, 0], color=colors[i])
"""------ACTUATED-------"""
for i in range(0, segs - 1):
ax.plot(points[0, i:i + 2],
points[1, i:i + 2],
color=colors[i])
else:
"""------NORMAL-------"""
''' PLOT UNACTUATED LINE HERE '''
# normal = np.zeros((segs+1))
# ax.plot(normal, color='grey',
# label="Initial pressure (P=P" + r'$_i$' + ")")
"""------ACTUATED-------"""
ax.plot(points[0, :],
points[1, :],
color='red',
label="Reduced-order model")
if any(value == True
for value in parameters.get('Curve fitting').values()):
m = points[0][-1] / (2 * pi)
if parameters.get('Curve fitting').get('Sin'):
curve = 20 * np.sin(points[0] / m)
elif parameters.get('Curve fitting').get('Cos'):
curve = 25 * np.cos(points[0] / m) - 25
elif parameters.get('Curve fitting').get('Custom'):
curve = []
func = parameters.get('Curve fitting').get('Custom func')
for _ in points[0] / m:
curve.append(eval(func))
ax.plot(points[0],
curve,
color='black',
alpha=0.85,
linestyle='--',
label='Desired profile')
if settings.get('Plot boundaries'):
to_tuple = [(x, y) for x, y in zip(limb.XY[0], limb.XY[1])]
line_check = LineString(to_tuple)
line_top = line_check.parallel_offset(1.9, side='left')
line_bottom = line_check.parallel_offset(1.9, side='right')
ax.plot(line_top.xy[0], line_top.xy[1])
ax.plot(line_bottom.xy[0], line_bottom.xy[1])
if settings.get('Overlay images'):
overlay_images(ax, limb)
diff = 20
x_min = min(points[0, :]) - diff
x_max = max(points[0, :]) + diff
y_min = min(points[1, :]) - diff
y_max = max(points[1, :]) + diff
''' THIS SETTING ALLOWS TO FOCUS THE WINDOW ON THE SELECTED SUBPLOT IF Top > 1 '''
# fig.canvas.mpl_connect('button_press_event', on_click)
plt.xlim(x_min, x_max)
plt.ylim(y_min, y_max)
plt.xlabel('X [mm]')
plt.ylabel('Y [mm]')
ax.xaxis.set_major_locator(MultipleLocator(20))
ax.grid(linestyle=':')
figManager = plt.get_current_fig_manager()
figManager.window.state('zoomed')
ax.legend(loc='lower center', bbox_to_anchor=(0.5, 1))
plt.show()
