class LassoSelectionMode(VispyMouseMode):
icon = 'glue_lasso'
tool_id = 'vispy:lasso'
action_text = 'Select data using a lasso selection'
def __init__(self, viewer):
super(LassoSelectionMode, self).__init__(viewer)
self.line = Line(color='purple',
width=2,
method='agg',
parent=self._vispy_widget.canvas.scene)
def activate(self):
self.reset()
def reset(self):
self.line_pos = []
self.line.set_data(np.zeros((0, 2), dtype=float))
self.line.parent = None
def press(self, event):
if event.button == 1:
self.line_pos.append(event.pos)
def move(self, event):
if event.button == 1 and event.is_dragging:
self.line_pos.append(event.pos)
self.line.set_data(np.array(self.line_pos, dtype=float))
self.line.parent = self._vispy_widget.canvas.scene
def release(self, event):
if event.button == 1:
if len(self.line_pos) > 0:
vx, vy = np.array(self.line_pos).transpose()
roi = Projected3dROI(roi_2d=PolygonalROI(vx, vy),
projection_matrix=self.projection_matrix)
self.apply_roi(roi)
self.reset()
self.viewer.toolbar.active_tool = None
# Here the mesh has a single boundary
open_mesh_boundary = stop.mesh_boundary(open_mesh)
print(open_mesh_boundary)
###############################################################################
# show the result
# WARNING : BrainObj should be added first before
visb_sc = splt.visbrain_plot(mesh=open_mesh, caption='open mesh')
# create points with vispy
for bound in open_mesh_boundary:
points = open_mesh.vertices[bound]
s_rad = SourceObj('rad', points, color='red', symbol='square',
radius_min=10)
visb_sc.add_to_subplot(s_rad)
lines = Line(pos=open_mesh.vertices[bound], width=10, color='b')
# wrap the vispy object using visbrain
l_obj = VispyObj('line', lines)
visb_sc.add_to_subplot(l_obj)
visb_sc.preview()
###############################################################################
# here is how to get the vertices that define the boundary of
# a texture on a mesh
# Let us first load example data
mesh = sio.load_mesh('../examples/data/example_mesh.gii')
# rotate the mesh for better visualization
transfo_flip = np.array([[-1, 0, 0, 0],[0, 1, 0, 0],[0, 0, -1, 0], [0, 0, 0, 1]])
mesh.apply_transform(transfo_flip)
# Load the example texture and compute its boundary
def __init__(self, parent):
super(SignalWidget, self).__init__(parent)
# Useful trnascripts
self.plugin = self.parent()
self.sd = self.plugin.sd
self.CONF_SECTION = self.parent().CONF_SECTION
# Variables
self.measurement_mode = False
self.curr_pc = None
self.sig_start = None
self.sig_stop = None
self.spect_type = 'spectrum' # spectrum, spectrogram
# General variables
self.low_lim = None
self.high_lim = None
# Sepctrum variables
self.mean_filter = None
# Setup camera
self.signal_camera = SignalCamera()
self.spectrum_camera = SignalCamera()
self.canvas = scene.SceneCanvas(show=True,
keys='interactive',
parent=self,
bgcolor=CONF.get(
self.CONF_SECTION, 'bgcolor'))
self.view_grid = self.canvas.central_widget.add_grid(margin=10)
# Signal
self.signal_view = self.view_grid.add_view(row=0,
col=1,
row_span=2,
camera=self.signal_camera)
axis_color = CONF.get(self.CONF_SECTION, 'axis_color')
self.signal_yaxis = AxisWidget(orientation='left',
axis_label='Amplitude',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.signal_yaxis.width_max = 60
self.view_grid.add_widget(self.signal_yaxis, row=0, col=0, row_span=2)
self.signal_xaxis = scene.AxisWidget(orientation='bottom',
axis_label='Time [s]',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.signal_xaxis.height_max = 55
self.view_grid.add_widget(self.signal_xaxis, row=2, col=1)
self.signal_yaxis.link_view(self.signal_view)
self.signal_xaxis.link_view(self.signal_view)
# Spectrum
self.spectrum_view = self.view_grid.add_view(
row=3, col=1, row_span=2, camera=self.spectrum_camera)
self.spectrum_yaxis = AxisWidget(orientation='left',
axis_label='Amplitude',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.spectrum_yaxis.width_max = 60
self.view_grid.add_widget(self.spectrum_yaxis,
row=3,
col=0,
row_span=2)
self.spectrum_xaxis = scene.AxisWidget(orientation='bottom',
axis_label='Frequency [Hz]',
axis_font_size=12,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.spectrum_xaxis.height_max = 55
self.view_grid.add_widget(self.spectrum_xaxis, row=5, col=1)
self.spectrum_yaxis.link_view(self.spectrum_view)
self.spectrum_xaxis.link_view(self.spectrum_view)
self.signal_line = Line(parent=self.signal_view.scene, width=1)
self.spectrum_line = Line(parent=self.spectrum_view.scene, width=1)
self.spectrogram = Spectrogram([0], parent=self.spectrum_view.scene)
# ----- Set layout -----
# Widget layout
layout = QVBoxLayout()
layout.setContentsMargins(0, 0, 0, 0)
layout.addWidget(self.canvas.native)
# Set the whole layout
self.setLayout(layout)
class SignalWidget(QWidget):
def __init__(self, parent):
super(SignalWidget, self).__init__(parent)
# Useful trnascripts
self.plugin = self.parent()
self.sd = self.plugin.sd
self.CONF_SECTION = self.parent().CONF_SECTION
# Variables
self.measurement_mode = False
self.curr_pc = None
self.sig_start = None
self.sig_stop = None
self.spect_type = 'spectrum' # spectrum, spectrogram
# General variables
self.low_lim = None
self.high_lim = None
# Sepctrum variables
self.mean_filter = None
# Setup camera
self.signal_camera = SignalCamera()
self.spectrum_camera = SignalCamera()
self.canvas = scene.SceneCanvas(show=True,
keys='interactive',
parent=self,
bgcolor=CONF.get(
self.CONF_SECTION, 'bgcolor'))
self.view_grid = self.canvas.central_widget.add_grid(margin=10)
# Signal
self.signal_view = self.view_grid.add_view(row=0,
col=1,
row_span=2,
camera=self.signal_camera)
axis_color = CONF.get(self.CONF_SECTION, 'axis_color')
self.signal_yaxis = AxisWidget(orientation='left',
axis_label='Amplitude',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.signal_yaxis.width_max = 60
self.view_grid.add_widget(self.signal_yaxis, row=0, col=0, row_span=2)
self.signal_xaxis = scene.AxisWidget(orientation='bottom',
axis_label='Time [s]',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.signal_xaxis.height_max = 55
self.view_grid.add_widget(self.signal_xaxis, row=2, col=1)
self.signal_yaxis.link_view(self.signal_view)
self.signal_xaxis.link_view(self.signal_view)
# Spectrum
self.spectrum_view = self.view_grid.add_view(
row=3, col=1, row_span=2, camera=self.spectrum_camera)
self.spectrum_yaxis = AxisWidget(orientation='left',
axis_label='Amplitude',
axis_font_size=12,
tick_label_margin=5,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.spectrum_yaxis.width_max = 60
self.view_grid.add_widget(self.spectrum_yaxis,
row=3,
col=0,
row_span=2)
self.spectrum_xaxis = scene.AxisWidget(orientation='bottom',
axis_label='Frequency [Hz]',
axis_font_size=12,
axis_color=axis_color,
tick_color=axis_color,
text_color=axis_color)
self.spectrum_xaxis.height_max = 55
self.view_grid.add_widget(self.spectrum_xaxis, row=5, col=1)
self.spectrum_yaxis.link_view(self.spectrum_view)
self.spectrum_xaxis.link_view(self.spectrum_view)
self.signal_line = Line(parent=self.signal_view.scene, width=1)
self.spectrum_line = Line(parent=self.spectrum_view.scene, width=1)
self.spectrogram = Spectrogram([0], parent=self.spectrum_view.scene)
# ----- Set layout -----
# Widget layout
layout = QVBoxLayout()
layout.setContentsMargins(0, 0, 0, 0)
layout.addWidget(self.canvas.native)
# Set the whole layout
self.setLayout(layout)
def set_spect_type(self, stype):
self.spect_type = stype
if stype == 'spectrum':
self.spectrum_line.visible = True
self.spectrogram.visible = False
self.spectrum_xaxis.axis.axis_label = 'Frequency [Hz]'
self.spectrum_yaxis.axis.axis_label = 'Amplitude'
elif stype == 'spectrogram':
self.spectrogram.visible = True
self.spectrum_line.visible = False
self.spectrum_xaxis.axis.axis_label = 'Time [s]'
self.spectrum_yaxis.axis.axis_label = 'Frequency [Hz]'
self.update_signals()
def recieve_input(self, event):
if event.type == 'key_press' and event.key == 'control':
self.measurement_mode = True
self.sig_start = None
self.sig_stop = None
elif event.type == 'key_release' and event.key == 'control':
self.measurement_mode = False
if event.type == 'mouse_press' and self.measurement_mode is True:
self.curr_pc = self.sd.curr_pc
rect_rel_w_pos = self.sd.rect_rel_w_pos
self.sig_start = int(rect_rel_w_pos * len(self.curr_pc.data))
self.spectrogram.fs = self.curr_pc.fsamp
if self.curr_pc is not None:
if self.curr_pc.fsamp != self.sd.curr_pc.fsamp:
self.low_lim = None
self.high_lim = None
self.curr_pc = self.sd.curr_pc
if self.low_lim is None:
self.plugin.tools_widget.general_tools \
.low_lim_le.setText('0')
self.plugin.tools_widget.general_tools \
.low_lim_le_validator.setRange(0, self.curr_pc.fsamp/2, 1)
if self.high_lim is None:
self.plugin.tools_widget.general_tools \
.high_lim_le.setText(str(self.curr_pc.fsamp/2))
self.plugin.tools_widget.general_tools \
.high_lim_le_validator.setRange(0, self.curr_pc.fsamp/2, 1)
if event.type == 'mouse_move' and self.measurement_mode is True:
self.sig_stop = int(self.sd.rect_rel_w_pos *
len(self.curr_pc.data))
self.update_signals()
def update_signals(self):
if self.sig_start is None or self.sig_stop is None:
return
if self.sig_start < self.sig_stop:
data = self.curr_pc.data[self.sig_start:self.sig_stop]
elif self.sig_start > self.sig_stop:
data = self.curr_pc.data[self.sig_stop:self.sig_start]
else:
return
if len(data) < 2:
return
# Signal line
s_x = 1 / self.curr_pc.fsamp
s_y = 1 / (np.max(data) - np.min(data))
s_z = 0
scale = [s_x, 1, 1]
# Translate
t_x = 0
t_y = -np.min(data)
t_z = 0
translate = [t_x, t_y, t_z]
transform = STTransform(scale, translate)
pos = np.c_[np.arange(len(data)), data]
self.signal_line.set_data(pos=pos, color=self.curr_pc.line_color)
self.signal_line.transform = transform
self.signal_camera.rect = (0, 0), (len(data) * s_x,
np.max(data) - np.min(data))
self.signal_camera.limit_rect = self.signal_camera.rect
if self.spect_type == 'spectrum':
s = np.abs(np.fft.fft(data))[:int(len(data) / 2)]
s[0] = 0
freqs = np.fft.fftfreq(data.size, 1 / self.curr_pc.fsamp)
freqs = freqs[:int(len(freqs) / 2)]
if self.mean_filter is not None and self.mean_filter > 1:
s = np.convolve(s,
np.ones(
(self.mean_filter, )) / self.mean_filter,
mode='valid')
if self.low_lim is not None:
res = np.where(freqs >= self.low_lim)[0]
if len(res) > 0:
low_lim_idx = res[0]
else:
low_lim_idx = 0
if self.high_lim is not None:
res = np.where(freqs <= self.high_lim)[0]
if len(res) > 0:
high_lim_idx = res[-1]
else:
high_lim_idx = len(freqs)
s_x = (self.curr_pc.fsamp / 2) / len(s)
s_y = 1
s_z = 1
scale = [s_x, s_y, s_z]
pos = np.c_[np.arange(len(s)), s]
self.spectrum_line.set_data(pos=pos, color=self.curr_pc.line_color)
self.spectrum_line.transform = STTransform(scale)
# Adjust camera limits
pos = (0, 0)
size = (freqs[-1], np.max(s))
self.spectrum_camera.limit_rect = pos, size
# Adjust camera view
freqs = freqs[low_lim_idx:high_lim_idx]
if len(freqs) == 0:
return
pos = (freqs[0], 0)
size = (freqs[-1] - freqs[0], np.max(s[low_lim_idx:high_lim_idx]))
self.spectrum_camera.rect = pos, size
elif self.spect_type == 'spectrogram':
self.spectrogram.x = data
freqs = self.spectrogram.freqs
if self.low_lim is not None:
res = np.where(freqs >= self.low_lim)[0]
if len(res) > 0:
low_lim_idx = res[0]
else:
low_lim_idx = 0
if self.high_lim is not None:
res = np.where(freqs <= self.high_lim)[0]
if len(res) > 0:
high_lim_idx = res[-1]
else:
high_lim_idx = len(freqs)
n_windows = (
(len(data) - self.spectrogram.n_fft) // self.spectrogram.step +
1)
if n_windows == 0 or len(freqs) == 0:
return
s_x = len(data) / self.curr_pc.fsamp
s_y = (self.curr_pc.fsamp / 2) / len(freqs)
s_z = 1
scale = [s_x, s_y, s_z]
self.spectrogram.transform = STTransform(scale)
# Adjust camera view
freqs = freqs[low_lim_idx:high_lim_idx - 1]
pos = (0, freqs[0])
size = (len(data) / self.curr_pc.fsamp, freqs[-1] - freqs[0])
self.spectrum_camera.rect = pos, size
# Adjust camera limits
self.spectrum_camera.limit_rect = pos, size
###############################################################################
# Generation of an open mesh
K = [-1, -1]
open_mesh = sps.generate_quadric(K, nstep=[5, 5])
open_mesh_boundary = stop.mesh_boundary(open_mesh)
# Visualization
visb_sc = splt.visbrain_plot(mesh=open_mesh, caption='open mesh')
for bound in open_mesh_boundary:
points = open_mesh.vertices[bound]
s_rad = SourceObj('rad',
points,
color='red',
symbol='square',
radius_min=10)
visb_sc.add_to_subplot(s_rad)
lines = Line(pos=open_mesh.vertices[bound], width=10, color='b')
# wrap the vispy object using visbrain
l_obj = VispyObj('line', lines)
visb_sc.add_to_subplot(l_obj)
visb_sc.preview()
###############################################################################
# Mapping onto a planar disk
disk_mesh = smap.disk_conformal_mapping(open_mesh)
# Visualization
visb_sc2 = splt.visbrain_plot(mesh=disk_mesh, caption='disk mesh')
for bound in open_mesh_boundary:
points = disk_mesh.vertices[bound]
s_rad = SourceObj('rad',
points,
color='red',
def __init__(self, parent):
super(SignalDisplay, self).__init__(parent)
# Covenience transcripts
self.main = self.parent()
# Widget behavior
self.setAcceptDrops(True)
# Plot variables
self.sample_map = []
self.plot_containers = []
# TODO: Selected signal plot used for data shifting, colors, etc
self.master_pc = None
self.master_plot = None # TODO - to be deleted
self.curr_pc = None
self.rect_rel_w_pos = None
self.rect_rel_h_pos = None
self.resize_flag = False
self.highlight_mode = False
self.measurement_mode = False
self.autoscale = False
self.disconts_processed = False
self.data_map = DataMap()
self.data_source = sm.ODS
self.data_array = None
# Widget layout
layout = QVBoxLayout()
layout.setContentsMargins(0, 0, 0, 0)
# These variables are assigned in channels plugin
self.hidden_channels = None
self.visible_channels = None
# Setup camera
self.camera = SignalCamera()
# Autoslide
self.slide_worker_stopped = True
# TODO: this should be i config
self.slide_worker = TimerWorker(1)
self.slide_worker_thread = QThread()
self.slide_worker.moveToThread(self.slide_worker_thread)
self.start_slide_worker.connect(self.slide_worker.run)
self.stop_slide_worker.connect(self.slide_worker.interupt)
self.slide_worker.time_passed.connect(self.autoslide)
self.slide_worker_thread.start()
# Vispy canvas
self.canvas = scene.SceneCanvas(show=True, keys='interactive',
parent=self,
bgcolor=CONF.get(self.CONF_SECTION,
'bgcolor'))
self.canvas.connect(self.on_key_press)
self.canvas.connect(self.on_key_release)
self.canvas.connect(self.on_mouse_move)
self.canvas.connect(self.on_mouse_press)
self.canvas.connect(self.on_mouse_release)
self.canvas.connect(self.on_mouse_wheel)
self.canvas.connect(self.on_resize)
# Timer to let the scene redraw if key is hit
self.event_time = time()
self.plot_update_done = False
# ??? Create two viewboxes - for labels and signals
self.signal_view = self.canvas.central_widget.add_view(
camera=self.camera)
self.cong_discontinuities = None
self.color_coding_mode = 0
self.color_palette = CONF.get(self.CONF_SECTION, 'color_palette')
self.update_cam_state()
# ----- Initial visuals operations-----
# TODO - Add crosshair color to CONF
# Measurements
ch_color = CONF.get(self.CONF_SECTION, 'init_crosshair_color')
self.crosshair = Crosshair(parent=self.signal_view.scene,
color=hex2rgba(ch_color))
m_color = CONF.get(self.CONF_SECTION, 'init_marker_color')
# TODO marker color
self.marker = Markers(parent=self.signal_view.scene)
self.xaxis = Axis(parent=self.signal_view.scene,
tick_direction=(0., 1.),
axis_width=1, tick_width=1,
anchors=('center', 'top'),
axis_color=m_color,
tick_color=m_color,
text_color=m_color)
self.x_tick_spacing = 1000
self.yaxis = Axis(parent=self.signal_view.scene,
tick_direction=(1., 0.),
axis_width=1, tick_width=1,
anchors=('left', 'center'),
axis_color=m_color,
tick_color=m_color,
text_color=m_color)
self.y_tick_spacing = 100
self.measure_line = Line(parent=self.signal_view.scene,
width=3, color=m_color)
# TODO - textbox
self.describe_text = Text(anchor_x='left',
anchor_y='bottom',
parent=self.signal_view.scene,
color=m_color)
# Signal highlighting
self.highlight_rec = Mesh(parent=self.signal_view.scene,
color=np.array([0., 0., 0., 0.]),
mode='triangle_fan')
# Grid
self.grid = None
# Discontinuity
self.disc_marker = LinearRegion(np.array([0, 0]),
np.array([[0., 0., 0., 0.],
[0., 0., 0., 0.]]),
parent=self.signal_view.scene)
self.signal_label_dict = {}
# Main signal visal with labels
w = CONF.get(self.CONF_SECTION, 'init_line_width')
self.signal_visual = Multiline(width=w,
parent=self.signal_view.scene)
self.label_visual = Text(anchor_x='left',
anchor_y='top',
parent=self.signal_view.scene)
# TODO - one set of x and y axes for measurements
# ----- Tool bar -----
btn_layout = QHBoxLayout()
for btn in self.setup_buttons():
if btn is None:
continue
btn.setAutoRaise(True)
btn.setIconSize(QSize(20, 20))
btn_layout.addWidget(btn)
# if options_button:
# btn_layout.addStretch()
# btn_layout.addWidget(options_button, Qt.AlignRight)
# TODO - this is temporary - solve the rendering in different thread
select_mode = QComboBox(self)
select_mode.insertItems(0, ['Browse', 'Research'])
antialias = CONF.get(self.CONF_SECTION, 'antialiasing')
if antialias == 'filter':
select_mode.setCurrentIndex(0)
elif antialias == 'min_max':
select_mode.setCurrentIndex(1)
select_mode.currentIndexChanged.connect(self.switch_display_mode)
btn_layout.addWidget(select_mode)
# Color coding
color_code = QComboBox(self)
color_code.insertItems(0, ['None', 'Channel', 'Group', 'Amplitude'])
color_code.currentIndexChanged.connect(self.switch_cc_mode)
btn_layout.addWidget(color_code)
# Metadata reload button
btn_layout.setAlignment(Qt.AlignLeft)
layout = create_plugin_layout(btn_layout)
# ----- Set layout -----
layout.addWidget(self.canvas.native)
# Set the whole layout
self.setLayout(layout)
# Connect signals
self.main.sig_file_opened.connect(self.initialize_data_map)
self.main.metadata_reloaded.connect(self.create_conglomerate_disconts)
self.plots_changed.connect(self.set_plot_update)
self.plots_changed.connect(self.subsample)
self.plots_changed.connect(self.rescale_grid)
self.input_recieved.connect(self.set_highlight_mode)
self.input_recieved.connect(self.show_measure_line)
self.canvas_resized.connect(self.update_subsample)
class SignalDisplay(QWidget):
# Attributes - tehcnically this is not a plugin but has the same attributes
CONF_SECTION = 'signal_display'
CONFIGWIDGET_CLASS = None
IMG_PATH = 'images'
DISABLE_ACTIONS_WHEN_HIDDEN = True
shortcut = None
# Signals
data_map_changed = pyqtSignal(DataMap, name='data_map_changed')
plots_changed = pyqtSignal(name='plots_changed')
# TODO: This will send a signal to event eveluator in the future
input_recieved = pyqtSignal(Event, name='input_recieved')
canvas_resized = pyqtSignal(name='canvas_resized')
subview_changed = pyqtSignal(name='subview_changed')
stop_slide_worker = pyqtSignal()
start_slide_worker = pyqtSignal()
def __init__(self, parent):
super(SignalDisplay, self).__init__(parent)
# Covenience transcripts
self.main = self.parent()
# Widget behavior
self.setAcceptDrops(True)
# Plot variables
self.sample_map = []
self.plot_containers = []
# TODO: Selected signal plot used for data shifting, colors, etc
self.master_pc = None
self.master_plot = None # TODO - to be deleted
self.curr_pc = None
self.rect_rel_w_pos = None
self.rect_rel_h_pos = None
self.resize_flag = False
self.highlight_mode = False
self.measurement_mode = False
self.autoscale = False
self.disconts_processed = False
self.data_map = DataMap()
self.data_source = sm.ODS
self.data_array = None
# Widget layout
layout = QVBoxLayout()
layout.setContentsMargins(0, 0, 0, 0)
# These variables are assigned in channels plugin
self.hidden_channels = None
self.visible_channels = None
# Setup camera
self.camera = SignalCamera()
# Autoslide
self.slide_worker_stopped = True
# TODO: this should be i config
self.slide_worker = TimerWorker(1)
self.slide_worker_thread = QThread()
self.slide_worker.moveToThread(self.slide_worker_thread)
self.start_slide_worker.connect(self.slide_worker.run)
self.stop_slide_worker.connect(self.slide_worker.interupt)
self.slide_worker.time_passed.connect(self.autoslide)
self.slide_worker_thread.start()
# Vispy canvas
self.canvas = scene.SceneCanvas(show=True, keys='interactive',
parent=self,
bgcolor=CONF.get(self.CONF_SECTION,
'bgcolor'))
self.canvas.connect(self.on_key_press)
self.canvas.connect(self.on_key_release)
self.canvas.connect(self.on_mouse_move)
self.canvas.connect(self.on_mouse_press)
self.canvas.connect(self.on_mouse_release)
self.canvas.connect(self.on_mouse_wheel)
self.canvas.connect(self.on_resize)
# Timer to let the scene redraw if key is hit
self.event_time = time()
self.plot_update_done = False
# ??? Create two viewboxes - for labels and signals
self.signal_view = self.canvas.central_widget.add_view(
camera=self.camera)
self.cong_discontinuities = None
self.color_coding_mode = 0
self.color_palette = CONF.get(self.CONF_SECTION, 'color_palette')
self.update_cam_state()
# ----- Initial visuals operations-----
# TODO - Add crosshair color to CONF
# Measurements
ch_color = CONF.get(self.CONF_SECTION, 'init_crosshair_color')
self.crosshair = Crosshair(parent=self.signal_view.scene,
color=hex2rgba(ch_color))
m_color = CONF.get(self.CONF_SECTION, 'init_marker_color')
# TODO marker color
self.marker = Markers(parent=self.signal_view.scene)
self.xaxis = Axis(parent=self.signal_view.scene,
tick_direction=(0., 1.),
axis_width=1, tick_width=1,
anchors=('center', 'top'),
axis_color=m_color,
tick_color=m_color,
text_color=m_color)
self.x_tick_spacing = 1000
self.yaxis = Axis(parent=self.signal_view.scene,
tick_direction=(1., 0.),
axis_width=1, tick_width=1,
anchors=('left', 'center'),
axis_color=m_color,
tick_color=m_color,
text_color=m_color)
self.y_tick_spacing = 100
self.measure_line = Line(parent=self.signal_view.scene,
width=3, color=m_color)
# TODO - textbox
self.describe_text = Text(anchor_x='left',
anchor_y='bottom',
parent=self.signal_view.scene,
color=m_color)
# Signal highlighting
self.highlight_rec = Mesh(parent=self.signal_view.scene,
color=np.array([0., 0., 0., 0.]),
mode='triangle_fan')
# Grid
self.grid = None
# Discontinuity
self.disc_marker = LinearRegion(np.array([0, 0]),
np.array([[0., 0., 0., 0.],
[0., 0., 0., 0.]]),
parent=self.signal_view.scene)
self.signal_label_dict = {}
# Main signal visal with labels
w = CONF.get(self.CONF_SECTION, 'init_line_width')
self.signal_visual = Multiline(width=w,
parent=self.signal_view.scene)
self.label_visual = Text(anchor_x='left',
anchor_y='top',
parent=self.signal_view.scene)
# TODO - one set of x and y axes for measurements
# ----- Tool bar -----
btn_layout = QHBoxLayout()
for btn in self.setup_buttons():
if btn is None:
continue
btn.setAutoRaise(True)
btn.setIconSize(QSize(20, 20))
btn_layout.addWidget(btn)
# if options_button:
# btn_layout.addStretch()
# btn_layout.addWidget(options_button, Qt.AlignRight)
# TODO - this is temporary - solve the rendering in different thread
select_mode = QComboBox(self)
select_mode.insertItems(0, ['Browse', 'Research'])
antialias = CONF.get(self.CONF_SECTION, 'antialiasing')
if antialias == 'filter':
select_mode.setCurrentIndex(0)
elif antialias == 'min_max':
select_mode.setCurrentIndex(1)
select_mode.currentIndexChanged.connect(self.switch_display_mode)
btn_layout.addWidget(select_mode)
# Color coding
color_code = QComboBox(self)
color_code.insertItems(0, ['None', 'Channel', 'Group', 'Amplitude'])
color_code.currentIndexChanged.connect(self.switch_cc_mode)
btn_layout.addWidget(color_code)
# Metadata reload button
btn_layout.setAlignment(Qt.AlignLeft)
layout = create_plugin_layout(btn_layout)
# ----- Set layout -----
layout.addWidget(self.canvas.native)
# Set the whole layout
self.setLayout(layout)
# Connect signals
self.main.sig_file_opened.connect(self.initialize_data_map)
self.main.metadata_reloaded.connect(self.create_conglomerate_disconts)
self.plots_changed.connect(self.set_plot_update)
self.plots_changed.connect(self.subsample)
self.plots_changed.connect(self.rescale_grid)
self.input_recieved.connect(self.set_highlight_mode)
self.input_recieved.connect(self.show_measure_line)
self.canvas_resized.connect(self.update_subsample)
# ----- Setup functions -----
def setup_buttons(self):
take_screenshot = create_toolbutton(self, icon='camera.svg',
tip='Take screenshot',
triggered=self.take_screenshot)
show_grid = create_toolbutton(self, icon='grid.svg',
tip='Show grid',
triggered=self.show_grid)
autoscale = create_toolbutton(self, icon='autoscale.svg',
tip='Autoscale',
triggered=self.set_autoscale)
save_data = create_toolbutton(self, icon='floppy-disk.svg',
tip='Save displayed data',
triggered=self.save_displayed_data)
# ---- TEMP !!! ------- -> move to the right side (add stretch)
reload_metadata = create_toolbutton(self, icon='reload.svg',
tip='Reload metadata',
toggled=self.main.
ss_reaload_worker)
autoslide = create_toolbutton(self, icon='play.svg',
tip='Autoslide',
toggled=self.ss_autoslide_worker)
# --------------
return (take_screenshot, show_grid, autoscale, save_data,
reload_metadata, autoslide)
# ----- Key bindings -----
def set_plot_update(self):
self.plot_update_done = True
def check_event_timer(self, time):
if time - self.event_time < .1:
return False
else:
return True
def set_event_timer(self):
self.event_time = time()
def on_key_press(self, event):
if event.handled:
return
modifiers = event.modifiers
plot_data_operators = ['up', 'down', 'left', 'right', 'q', 'a']
# TODO: there should be a key mapper in the future! - python dictionary
if event.type == 'key_press':
if (event.key in plot_data_operators
and self.plot_update_done
and self.check_event_timer(time())):
if event.key == 'Up':
self.scale_plot_data(True)
if event.key == 'Down':
self.scale_plot_data(False)
# These operations require data pull, introduced a timer
# NOTE: we now know that rendering of big data takes time
if event.key == 'Left':
self.plot_update_done = False
if 'shift' in modifiers: # Partial shift
self.shift_plot_data(False, 0.5)
else:
self.shift_plot_data(False)
if event.key == 'Right':
self.plot_update_done = False
if 'shift' in modifiers: # Partial shift
self.shift_plot_data(True, 0.5)
else:
self.shift_plot_data(True)
if event.key == 'q':
self.plot_update_done = False
self.change_time_span(True)
if event.key == 'a':
self.plot_update_done = False
self.change_time_span(False)
self.set_event_timer()
event.handled = True
else:
self.input_recieved.emit(event)
else:
event.handled = False
def on_key_release(self, event):
if event.handled:
return
if event.type == 'key_release':
self.input_recieved.emit(event)
event.handled = True
else:
event.handled = False
# ----- Highlight / measurement mode , canvas behavior -----
def set_highlight_mode(self, event):
if event.type not in ('key_press', 'key_release'):
return
if event.key not in ('shift', 'control'):
return
if event.type == 'key_press' and event.key == 'shift':
self.highlight_mode = True
self.highlight_rec.visible = True
elif event.type == 'key_press' and event.key == 'control':
self.measurement_mode = True
self.crosshair.visible = True
self.marker.visible = True
self.xaxis.visible = True
self.yaxis.visible = True
elif event.type == 'key_release' and event.key == 'shift':
self.highlight_mode = False
self.highlight_rec.visible = False
elif event.type == 'key_release' and event.key == 'control':
self.measurement_mode = False
self.crosshair.visible = False
self.marker.visible = False
self.xaxis.visible = False
self.yaxis.visible = False
self.measure_line.visible = False
self.describe_text.visible = False
def on_mouse_move(self, event):
if 1 in event.buttons or 2 in event.buttons and not event.modifiers:
self.subview_changed.emit()
# Get position relative to zoom
pos = event.pos[:2]
w = self.signal_view.width
h = self.signal_view.height
rel_w_pos = pos[0] / w
# TODO: flip Vispy axis
rel_h_pos = (h-pos[1]) / h
rect = self.camera.rect
self.rect_rel_w_pos = rect.left + (rel_w_pos * rect.width)
self.rect_rel_h_pos = rect.bottom + (rel_h_pos * rect.height)
# Determine the signal plot
rows = self.visible_channels.get_row_count()
cols = self.visible_channels.get_col_count()
sig_w_pos = self.rect_rel_w_pos * cols
sig_h_pos = self.rect_rel_h_pos * rows
for pc in self.get_plot_containers():
if ((pc.plot_position[0]
< sig_w_pos
< pc.plot_position[0]+1)
and (pc.plot_position[1]
< sig_h_pos
< pc.plot_position[1]+1)):
self.curr_pc = pc
break
# ??? Instead of modes use event.modifiers???
if self.highlight_mode:
self.highlight_signal(self.curr_pc)
if self.measurement_mode:
self.crosshair.set_data([self.rect_rel_w_pos,
self.rect_rel_h_pos])
n_channels = self.visible_channels.get_row_count()
# Get the location of data point
s_y = self.curr_pc.ufact*self.curr_pc.scale_factor
t_y = ((-np.nanmean(self.curr_pc.data)
* self.curr_pc.ufact
* self.curr_pc.scale_factor)
+ ((0.5+self.curr_pc.plot_position[1]) / n_channels))
data_pos = self.curr_pc.data[int(self.rect_rel_w_pos
* len(self.curr_pc.data))]
data_pos *= s_y
data_pos += t_y
self.marker.set_data(np.array([[self.rect_rel_w_pos, data_pos]]))
# TODO: determine margins
# Axes
t_y = (self.curr_pc.plot_position[1] / n_channels)
y_margin = 0
self.xaxis.pos = [[rect.left,
t_y + y_margin],
[rect.left+(rect.width*self.x_tick_spacing),
t_y + y_margin]]
rel_diff = (rect.right - rect.left) * np.diff(pc.uutc_ss)
self.xaxis.domain = tuple([0, rel_diff/1000000])
s = [1/self.x_tick_spacing, 1]
t = [rect.left-rect.left*s[0], 0]
self.xaxis.transform = scene.transforms.STTransform(s, t)
x_margin = 0
self.yaxis.pos = [[rect.left + x_margin,
t_y],
[rect.left + x_margin,
t_y + ((1/n_channels)*self.y_tick_spacing)]]
s = [1, 1/self.y_tick_spacing]
t = [0, t_y-t_y*s[1]]
self.yaxis.transform = scene.transforms.STTransform(s, t)
lpos = self.measure_line.pos
if lpos is not None:
fixed = lpos[0]
right_angle = np.array([fixed[0], data_pos])
moving = np.array([self.rect_rel_w_pos, data_pos])
whole_line = np.vstack([fixed, right_angle, moving])
self.measure_line.set_data(pos=whole_line)
# Time
max_step = 1/self.curr_pc.fsamp
time_dist = moving[0]-fixed[0]
time_dist *= np.diff(self.curr_pc.uutc_ss)[0] / 1e6
time_dist -= time_dist % max_step
oround = int(np.ceil((np.log10(self.curr_pc.fsamp))))
time_str = format(time_dist, '.'+str(oround)+'f')+' s'
time_str_pos = moving.copy()
# Amplitude
max_step = self.curr_pc.ufact
amp_dist = (moving[1] - fixed[1]) / s_y
amp_dist *= max_step
amp_dist -= amp_dist % amp_dist
amp_str = (format(amp_dist, '.5f') + ' ' + self.curr_pc.unit)
amp_str_pos = moving.copy()
amp_str_pos[0] = moving[0]
fsize = self.describe_text.font_size
amp_str_pos[1] += (((fsize+1)*rect.height)
/ self.signal_view.height)
self.describe_text.text = [time_str, amp_str]
self.describe_text.pos = [time_str_pos, amp_str_pos]
self.describe_text.color = np.array([[1., 1., 1., 1.],
[1., 1., 1., 1.]],
dtype=np.float32)
self.input_recieved.emit(event)
def show_measure_line(self, event):
if event.type != 'mouse_press':
return
modifiers = event.modifiers
if 'control' in modifiers:
# Get position relative to zoom
self.measure_line.visible = True
pos = self.marker._data['a_position'][0][:2]
self.measure_line.set_data(pos=np.tile(pos, 3).reshape([3, 2]))
self.describe_text.visible = True
def on_mouse_press(self, event):
self.input_recieved.emit(event)
def on_mouse_release(self, event):
self.subsample()
self.update_labels()
def on_mouse_wheel(self, event):
# TODO: subsample for zoom
# Get x_pos
x_pos = event.pos[0]
# Get the zoomed area
def on_resize(self, event):
self.resize_flag = True
if np.any(self.main.signal_display.data_map['ch_set']):
self.canvas_resized.emit()
# ----- Screenshot -----
def save_image(self):
im = pil_Image.fromarray(self.screen_img)
save_dialog = QFileDialog(self)
save_dialog.selectFile('pysigview_screenshot.png')
save_path = save_dialog.getSaveFileName(self, 'Save File',
get_home_dir(),
"Images (*.png *.tiff *.jpg)")
path = save_path[0]
if not any([x for x in ['.png', '.tiff', '.jpg'] if x in path]):
path += '.png'
im.save(path)
return
def close_screenshot(self):
self.screenshot_diag.reject()
def take_screenshot(self):
self.screen_img = self.canvas.render()
# Pop the screenshot window
self.screenshot_diag = QDialog(self)
self.screenshot_diag.setModal(False)
# Set the image
canvas = scene.SceneCanvas(show=True, size=self.canvas.size,
parent=self.screenshot_diag)
view = canvas.central_widget.add_view()
Image(self.screen_img, parent=view.scene)
layout = QVBoxLayout()
layout.setContentsMargins(0, 0, 0, 0)
layout.addWidget(canvas.native)
button_layout = QHBoxLayout()
cancel_btn = QPushButton('Cancel')
cancel_btn.clicked.connect(self.close_screenshot)
button_layout.addWidget(cancel_btn)
save_btn = QPushButton('Save')
save_btn.clicked.connect(self.save_image)
button_layout.addWidget(save_btn)
layout.addLayout(button_layout)
self.screenshot_diag.setLayout(layout)
self.screenshot_diag.setVisible(True)
# ----- Grid -----
def show_grid(self):
if not len(self.get_plot_containers()):
return
if self.grid is None:
rows = self.visible_channels.get_row_count()
if rows:
y_scale = 1/rows
else:
y_scale = 1
if self.master_plot:
pass # TODO
else:
uutc_ss = self.data_map.get_active_largest_ss()
span_secs = np.diff(uutc_ss) / 1e6
order_m = 10 ** (np.floor(np.log10(span_secs)))
x_scale = order_m / span_secs
c = CONF.get(self.CONF_SECTION, 'grid_color')
self.grid = GridLines(scale=(x_scale, y_scale),
color=c,
parent=self.signal_view.scene)
else:
self.grid.parent = None
self.grid = None
def rescale_grid(self):
if self.grid is not None:
self.grid.parent = None
self.grid = None
rows = self.visible_channels.get_row_count()
if rows:
y_scale = 1/rows
else:
y_scale = 1
if self.master_plot:
pass # TODO
else:
uutc_ss = self.data_map.get_active_largest_ss()
span_secs = np.diff(uutc_ss) / 1e6
order_m = 10 ** (np.floor(np.log10(span_secs)))
x_scale = order_m / span_secs
self.grid = GridLines(scale=(x_scale, y_scale),
parent=self.signal_view.scene)
# ----- Autoscale -----
def set_autoscale(self):
pcs = self.get_plot_containers()
for pc in pcs:
if pc.autoscale:
pc.autoscale = False
pc.scale_factor = 1
else:
pc.autoscale = True
self.update_signals()
# ----- Save displayed data -----
def save_displayed_data(self):
# Bring up save dialog
save_dialog = QFileDialog(self)
save_dialog.setDefaultSuffix(".pkl")
file_types = "Python pickle (*.pkl);;Matlab (*.mat)"
save_path = save_dialog.getSaveFileName(self, 'Save displayed data',
get_home_dir(),
file_types)
# Get the data from visuals
names = []
data = []
for pc in self.get_plot_containers():
names.append(pc.container.item_widget.label.text())
data.append(self.data_array[pc.data_array_pos][0])
data = np.vstack(data)
dict_out = {'channel_names': names,
'data': data}
path = save_path[0]
# Evaluate the extension and save
if save_path[1] == "Python pickle (*.pkl)":
if '.pkl' not in path:
path += '.pkl'
with open(path, 'wb') as fid:
pickle.dump(dict_out, fid)
elif save_path[1] == "Matlab (*.mat)":
if '.mat' not in path:
path += '.mat'
savemat(path, dict_out)
return
# ----- Display mode switch -----
def switch_display_mode(self, mode):
if mode == 0:
CONF.set(self.CONF_SECTION, 'antialiasing', 'filter')
elif mode == 1:
CONF.set(self.CONF_SECTION, 'antialiasing', 'min_max')
self.update_subsample()
if len(self.data_map.get_active_channels()):
self.set_plot_data()
return
def update_subsample(self):
antialias = CONF.get(self.CONF_SECTION, 'antialiasing')
pcs = self.get_plot_containers()
for pc in pcs:
if antialias == 'filter':
pc.N = int(self.canvas.central_widget.width)
elif antialias == 'min_max':
pc.N = None
# ----- Color coding -----
def switch_cc_mode(self, coding):
self.color_coding_mode = coding
self.color_code()
def color_code(self):
pcs = self.get_plot_containers()
if not len(pcs):
return
if self.color_coding_mode == 0:
c = hex2rgba(CONF.get(self.CONF_SECTION, 'init_line_color'))
for pc in pcs:
pc.line_color = c
pc.container.item_widget.color_select.set_color(c)
self.update_labels()
elif self.color_coding_mode == 1:
# Channels
#TODO - in prefs, color.get_colormaps()
cm = color.get_colormap(self.color_palette)
colors = cm[np.linspace(0, 1, len(pcs))]
for pc, c in zip(pcs, colors):
pc.line_color = c.rgba[0]
pc.container.item_widget.color_select.set_color(c.rgba[0])
self.update_labels()
# Acquire the colors based on number of channels
# ???Introduce a limit??? If not the channels might be too simliar
elif self.color_coding_mode == 2:
# Groups
ch_names = [x.orig_channel for x in pcs]
g_names = []
for ch_name in ch_names:
stub = ''.join([i for i in ch_name if not i.isdigit()])
g_names.append(stub)
g_names = list(set(g_names))
# TODO - in prefs, color.get_colormaps()
cm = color.get_colormap(self.color_palette)
colors = cm[np.linspace(0, 1, len(g_names))]
for pc in pcs:
stub = ''.join([i for i in pc.orig_channel if not i.isdigit()])
c = colors[g_names.index(stub)]
pc.line_color = c.rgba[0]
pc.container.item_widget.color_select.set_color(c.rgba[0])
self.update_labels()
elif self.color_coding_mode == 3:
# Amplitude
pass
else:
pass
self.update_signals()
# ----- Autoslide -----
def autoslide(self):
self.shift_plot_data(True)
def ss_autoslide_worker(self):
if self.slide_worker_stopped:
self.start_slide_worker.emit()
self.slide_worker_stopped = False
else:
self.stop_slide_worker.emit()
self.slide_worker_stopped = True
# ----- Camera handling -----
def reset_cam(self):
self.camera.set_state(self.orig_cam_state)
def update_cam_state(self):
self.orig_cam_state = self.camera.get_state()
# ----- QT Widget behavior (drag/drops) -----
def dragEnterEvent(self, event):
if event.mimeData().hasUrls:
event.accept()
else:
event.ignore()
def dropEvent(self, event):
event.setDropAction(Qt.CopyAction)
source = event.source()
# The source of drop fork
if event.source() is None: # Dropped from outside of Qt app
event.accept()
url = event.mimeData().urls()[0]
self.main.open_data_source(url.toLocalFile())
elif event.source() == self.hidden_channels: # Hidden channel list
event.accept()
self.visible_channels.insert_items([x.text() for x
in source.drag_items])
else:
event.ignore()
def enterEvent(self, event):
self.canvas.native.setFocus()
# ----- Data handling -----
def get_minmax_idxs(self, sig, step):
N = int(np.ceil(len(sig) / step))
max_idx = np.empty(N, 'uint32')
min_idx = np.empty(N, 'uint32')
trans_sig = np.resize(sig, (N, step))
max_idx = trans_sig.argmax(1)
max_idx += (np.arange(N)*step)
max_idx[max_idx >= len(sig)] = len(sig) - 1
min_idx = trans_sig.argmin(1)
min_idx += (np.arange(N)*step)
min_idx[min_idx >= len(sig)] = len(sig) - 1
return max_idx, min_idx
def subsample(self):
rect = self.signal_view.camera.rect
w = rect.width
max_viewed_points = 10000
right = rect.right
left = rect.left
if left < 0:
left = 0
if right > 1:
right = 1
vis_pos = self.signal_visual.pos
vis_index = self.signal_visual.index
indices = np.hstack([0, np.cumsum([len(x) for x in vis_pos])[:-1]])
for pc in self.get_plot_containers():
pos_i = pc._visual_array_idx
line_len = len(vis_pos[pos_i])
fraction = int((line_len * w) // max_viewed_points)
li = int(np.floor(left*line_len))
ri = int(np.floor(right*line_len))
if fraction <= 1:
vis_index[indices[pos_i]:indices[pos_i]+line_len] = 0
vis_index[li + indices[pos_i]:ri + indices[pos_i]] = 1
else:
max_idx, min_idx = self.get_minmax_idxs(vis_pos[pos_i][li:ri],
fraction)
# Corect the max min indices to line pos
max_idx += (li + indices[pos_i])
min_idx += (li + indices[pos_i])
vis_index[indices[pos_i]:indices[pos_i]+line_len] = 0
vis_index[max_idx] = 1
vis_index[min_idx] = 1
self.signal_visual.index = vis_index
# ----- Discontinuities -----
def create_conglomerate_disconts(self):
# TODO: what if the channels are changed? This should not be run!
disconts = sm.ODS.data_map['discontinuities']
chan_mask = self.main.signal_display.data_map['ch_set']
disconts = disconts[chan_mask]
# Do not process disconts that are already processed
if self.disconts_processed:
proc_start = self.cong_discontinuities[0][0]
proc_stop = self.cong_discontinuities[-1][-1]
for i in range(len(disconts)):
disconts[i] = disconts[i][~((disconts[i][:, 0] >= proc_start)
& (disconts[i][:, 1] <= proc_stop))]
# Channel with the fewest disconts
min_disc_ch_idx = np.argmin([len(x) for x in disconts])
min_disc_ch = disconts[min_disc_ch_idx]
conglom_discs = []
# Iterate over the rest of channels
for di, intersect in enumerate(min_disc_ch):
ch_count = 1
for cdi, ch_discs in enumerate(np.delete(disconts,
min_disc_ch_idx)):
# Get starts in inclusive intervals
idx = np.where((intersect[0] <= ch_discs[:, 0])
& (ch_discs[:, 0] <= intersect[1]))[0]
if len(idx):
intersect[0] = ch_discs[idx][0][0]
# Get stops in inclusive intervals
idx = np.where((intersect[0] <= ch_discs[:, 1])
& (ch_discs[:, 1] <= intersect[1]))[0]
if len(idx):
intersect[1] = ch_discs[idx][0][1]
ch_count += 1
# Is the discontinuity in all channels?
if ch_count == len(disconts):
conglom_discs.append(intersect)
if self.disconts_processed:
if len(conglom_discs):
self.cong_discontinuities = np.concatenate(
[self.cong_discontinuities, np.array(conglom_discs)])
else:
self.cong_discontinuities = np.array(conglom_discs)
self.disconts_processed = True
# ----- Data map operations -----
def initialize_data_map(self):
self.data_map.setup_data_map(sm.ODS.data_map._map)
self.data_map.reset_data_map()
# TODO: what if there are two channels with the same orig_channels
def update_data_map_channels(self):
"""
Updates data_map from visible_channels pane and reloads the data.
"""
pcs = self.get_plot_containers()
# Check if some channels are duplicate
channels = []
uutc_ss = []
for pc in pcs:
pc_chans = [pc.orig_channel] + pc.add_channels
# Get the channels in the current plot container
for o_ch in pc_chans:
# Check if the channel is already in the list
if o_ch in channels:
ch_idx = channels.index(o_ch)
if pc.uutc_ss[0] < uutc_ss[ch_idx][0]:
uutc_ss[ch_idx][0] = pc.uutc_ss[0]
if pc.uutc_ss[1] > uutc_ss[ch_idx][1]:
uutc_ss[ch_idx][1] = pc.uutc_ss[1]
else:
channels.append(o_ch)
uutc_ss.append(pc.uutc_ss)
self.data_map.reset_data_map()
if len(channels) == 0:
return
self.data_map.set_channel(np.array(channels),
np.array(uutc_ss))
self.create_conglomerate_disconts()
self.check_data_map_uutc_ss()
# Reset data_array where the channel is not set
if self.data_array is not None:
self.data_array[
~self.data_map._map['ch_set']] = np.array(0, 'float32')
# ----- Load data start -----
def get_plot_containers(self):
items = self.visible_channels.get_container_items()
return [x.pvc for x in items]
# XXX - this could probably be made into a general plot_container function
def add_signal_container(self, orig_channel):
container_items = self.visible_channels.get_container_items()
mf_scale_fatcor = None
scale_factors = [x.pvc.scale_factor for x in container_items]
if scale_factors:
# In case of half/half, let python choose :-)
mf_scale_fatcor = max(scale_factors, key=scale_factors.count)
# Get the max span and assign to new signals
largest_ss = self.data_map.get_active_largest_ss()
pc = SignalContainer(orig_channel)
ci = sm.ODS.data_map['channels'] == pc.orig_channel
ci_entry = sm.ODS.data_map[ci]
pc.fsamp = ci_entry['fsamp'][0]
pc.unit = ci_entry['unit'][0]
pc.ufact = ci_entry['ufact'][0]
pc.nsamp = ci_entry['nsamp'][0]
pc.start_time = ci_entry['uutc_ss'][0][0]
antialias = CONF.get(self.CONF_SECTION, 'antialiasing')
if antialias == 'filter':
pc.N = int(self.canvas.central_widget.width)
elif antialias == 'min_max':
pc.N = None
c = hex2rgba(CONF.get(self.CONF_SECTION, 'init_line_color'))
pc.line_color = np.array(c)
pc.data_array_pos = [np.where(ci)[0][0]]
# Scale factor
if mf_scale_fatcor:
pc.scale_factor = mf_scale_fatcor
# Time span
init_tscale = CONF.get(self.CONF_SECTION, 'init_time_scale')*1e6
if np.diff(largest_ss):
pc.uutc_ss = largest_ss
else:
pc.uutc_ss = [pc.start_time, pc.start_time+init_tscale]
return pc
def side_flash(self, color=None):
color = hex2rgba(CONF.get(self.CONF_SECTION, 'side_flash_color'))
if self.discont_side == 1: # left
pos = np.array([0., 0.1])
color = np.vstack([color,
[0., 0., 0., 0.]])
elif self.discont_side == 2: # right
pos = np.array([0.9, 1.0])
color = np.vstack([[0., 0., 0., 0.],
color])
else:
pos = np.array([0., 0.])
color = np.zeros([2, 4])
self.disc_marker.set_data(pos, color)
return
def check_uutc_ss(self, uutc_ss):
# Checks recording start and stop
span = np.diff(uutc_ss)[0]
if uutc_ss[0] < sm.ODS.recording_info['recording_start']:
uutc_ss[0] = sm.ODS.recording_info['recording_start']
if span < sm.ODS.recording_info['recording_duration']:
uutc_ss[1] = uutc_ss[0] + span
if uutc_ss[1] > sm.ODS.recording_info['recording_end']:
uutc_ss[1] = sm.ODS.recording_info['recording_end']
if span < sm.ODS.recording_info['recording_duration']:
uutc_ss[0] = uutc_ss[1] - span
self.discont_side = 0
# Checks for discontinuities
if self.cong_discontinuities is not None:
max_span = np.diff(self.data_map.get_active_largest_ss())
large_disconts_idxs = np.diff(self.cong_discontinuities) > max_span
large_disconts_idxs = large_disconts_idxs.ravel()
large_disconts = self.cong_discontinuities[large_disconts_idxs]
starts = large_disconts[:, 0] <= uutc_ss[0]
stops = uutc_ss[0] <= large_disconts[:, 1]
in_discont_idx = np.where(starts & stops)[0]
if len(in_discont_idx):
in_discont = large_disconts[in_discont_idx][0]
uutc_ss[0] = in_discont[1]
uutc_ss[1] = uutc_ss[0] + span
self.discont_side = 1 # left side
self.curr_discont = in_discont
starts = large_disconts[:, 0] <= uutc_ss[1]
stops = uutc_ss[1] <= large_disconts[:, 1]
in_discont_idx = np.where(starts & stops)[0]
if len(in_discont_idx):
in_discont = large_disconts[in_discont_idx][0]
uutc_ss[1] = in_discont[0]
uutc_ss[0] = uutc_ss[1] - span
self.discont_side = 2 # right side
self.curr_discont = in_discont
return uutc_ss
def check_data_map_uutc_ss(self):
corrected_uutc_ss = []
for uutc_ss in self.data_map.get_active_uutc_ss():
corrected_uutc_ss.append(self.check_uutc_ss(uutc_ss))
channels = self.data_map.get_active_channels()
self.data_map.set_channel(channels, corrected_uutc_ss)
self.data_map_changed.emit(self.data_map)
self.side_flash()
def check_pcs_uutc_ss(self):
for pc in self.get_plot_containers():
pc.uutc_ss = self.check_uutc_ss(pc.uutc_ss)
def calculate_sample(self, pc):
ch_i = np.where(self.data_map['channels'] == pc.orig_channel)[0]
dm_uutc_ss = self.data_map['uutc_ss'][ch_i][0]
start = int(((pc.uutc_ss[0] - dm_uutc_ss[0]) / 1e6) * pc.fsamp)
stop = int(((pc.uutc_ss[1] - dm_uutc_ss[0]) / 1e6) * pc.fsamp)
return start, stop
# TODO - when chnaging individual channel time scale
# the set_plot_data function is called twice - eliminate
def set_plot_data(self, uutc_ss=None, channels=None):
if self.data_array is None:
first_load = True
else:
first_load = False
# This check whether provider data source is a buffer
if getattr(sm.PDS, "is_available", None):
while not sm.PDS.is_available(self.data_map):
sleep(0.1)
continue
if len(self.data_map.get_active_channels()) == 0:
return
self.data_array = sm.PDS.get_data(self.data_map)
pcs = self.get_plot_containers()
for pc in pcs:
start, stop = self.calculate_sample(pc)
pc.data = np.array([x[start:stop] for x
in self.data_array[pc.data_array_pos]])
if first_load:
self.autoscale_plot_data(pcs[0])
for pc in pcs[1:]:
pc.scale_factor = pcs[0].scale_factor
self.update_signals()
if self.resize_flag:
self.resize_flag = False
return
# ----- Signal updating functions -----
def update_labels(self):
"""
Update names, positions and labels
"""
# Get current view left boundry
rect = self.signal_view.camera.rect
left = rect.left
name_list = []
pos_list = []
color_list = []
for pc in self.get_plot_containers():
if self.signal_visual.visibility[pc._visual_array_idx]:
name_list.append(pc.name)
color_list.append(pc.line_color)
# Label position
l_x = pc.plot_position[0] + left
l_y = (pc.plot_position[1]
/ self.visible_channels.get_row_count())
l_y += 1 / self.visible_channels.get_row_count()
y_shift = (pc.plot_position[2]
/ self.canvas.central_widget.height)
l_y -= y_shift * self.label_visual.font_size
pos_list.append([l_x, l_y, 0])
if len(name_list) == 0:
self.label_visual.text = None
else:
self.label_visual.text = name_list
self.label_visual.pos = pos_list
self.label_visual.color = np.c_[color_list]
def update_signals(self):
scales = []
offsets = []
color_list = []
data = np.empty(len(self.get_plot_containers()), object)
visibility = []
for li, pc in enumerate(self.get_plot_containers()):
data[li] = pc.data
pc._visual_array_idx = li
visibility.append(pc.visible)
if pc.autoscale:
self.autoscale_plot_data(pc)
# Scale
s_x = 1/len(pc.data)
s_y = pc.ufact*pc.scale_factor
s_z = 0
scales.append([s_x, s_y, s_z])
# Translate
t_x = pc.plot_position[0]
t_y = ((-np.nanmean(pc.data)
* pc.ufact
* pc.scale_factor)
+ ((0.5+pc.plot_position[1])
/ self.visible_channels.get_row_count()))
t_z = pc.plot_position[2]
offsets.append([t_x, t_y, t_z])
color_list.append(pc.line_color)
if len(data) == 0:
pos = np.empty(1, dtype=object)
pos[0] = np.array([0], dtype=np.float32)
self.signal_visual.pos = pos
else:
self.signal_visual.set_data(pos=data,
scales=scales, offsets=offsets,
color=color_list,
visibility=visibility)
self.update_labels()
self.plots_changed.emit()
def move_to_time(self, midpoint):
"""
Moves the view to requested time(will be in the middle)
"""
a_channels = self.data_map.get_active_channels()
a_uutc_ss = self.data_map.get_active_uutc_ss()
a_spans = np.diff(a_uutc_ss).ravel()
a_uutc_ss[:, 0] = midpoint - (a_spans / 2)
a_uutc_ss[:, 1] = midpoint + (a_spans / 2)
self.data_map.set_channel(a_channels, a_uutc_ss)
# Update individual plot containers
for pc in self.main.signal_display.get_plot_containers():
span = np.diff(pc.uutc_ss)[0]
pc.uutc_ss[0] = int(midpoint - (span / 2))
pc.uutc_ss[1] = int(midpoint + (span / 2))
self.check_pcs_uutc_ss()
self.check_data_map_uutc_ss()
self.set_plot_data()
def shift_plot_data(self, forward=True, shift_span=None):
# Take care of discontinuities
if self.discont_side:
span = np.diff(self.data_map.get_active_largest_ss())[0]
if self.discont_side == 1 and not forward:
midpoint_to_go = self.curr_discont[0] - (span / 2)
self.move_to_time(midpoint_to_go)
return
elif self.discont_side == 2 and forward:
midpoint_to_go = self.curr_discont[1] + (span / 2)
self.move_to_time(midpoint_to_go)
return
if self.master_plot:
uutc_ss = self.master_plot.uutc_ss
base_span = np.diff(uutc_ss)
else:
base_span = np.diff(self.data_map.get_active_largest_ss())[0]
if shift_span is None:
span = base_span
elif shift_span < 1:
span = base_span * shift_span
else:
span = shift_span
a_channels = self.data_map.get_active_channels()
a_uutc_ss = self.data_map.get_active_uutc_ss()
if forward:
a_uutc_ss += int(span)
self.data_map.set_channel(a_channels, a_uutc_ss)
else:
a_uutc_ss -= int(span)
self.data_map.set_channel(a_channels, a_uutc_ss)
for pc in self.get_plot_containers():
if forward:
pc.uutc_ss[0] += int(span)
pc.uutc_ss[1] += int(span)
else:
pc.uutc_ss[0] -= int(span)
pc.uutc_ss[1] -= int(span)
self.check_pcs_uutc_ss()
self.check_data_map_uutc_ss()
self.set_plot_data()
return
def change_time_span(self, up=True, channels=None, scale=2):
"""
Changes the time scale of plots
"""
a_channels = self.data_map.get_active_channels()
a_uutc_ss = self.data_map.get_active_uutc_ss()
a_spans = np.diff(a_uutc_ss).ravel()
a_midpoints = np.sum(a_uutc_ss, 1) / 2
if up:
a_uutc_ss[:, 0] = a_midpoints - (scale * (a_spans / 2))
a_uutc_ss[:, 1] = a_midpoints + (scale * (a_spans / 2))
self.data_map.set_channel(a_channels, a_uutc_ss)
else:
a_uutc_ss[:, 0] = a_midpoints - ((1/scale) * (a_spans / 2))
a_uutc_ss[:, 1] = a_midpoints + ((1/scale) * (a_spans / 2))
self.data_map.set_channel(a_channels, a_uutc_ss)
for pc in self.get_plot_containers():
span = np.diff(pc.uutc_ss)[0]
midpoint = np.sum(pc.uutc_ss) / 2
if up:
pc.uutc_ss[0] = midpoint - (scale * (span / 2))
pc.uutc_ss[1] = midpoint + (scale * (span / 2))
else:
pc.uutc_ss[0] = midpoint - ((1/scale) * (span / 2))
pc.uutc_ss[1] = midpoint + ((1/scale) * (span / 2))
self.check_pcs_uutc_ss()
self.check_data_map_uutc_ss()
self.set_plot_data()
return
def set_time_span_all(self, span):
a_channels = self.data_map.get_active_channels()
a_uutc_ss = self.data_map.get_active_uutc_ss()
a_midpoints = np.sum(a_uutc_ss, 1) / 2
a_uutc_ss[:, 0] = a_midpoints - (span / 2)
a_uutc_ss[:, 1] = a_midpoints + (span / 2)
self.data_map.set_channel(a_channels, a_uutc_ss)
for pc in self.get_plot_containers():
midpoint = np.sum(pc.uutc_ss) / 2
pc.uutc_ss[0] = midpoint - (span / 2)
pc.uutc_ss[1] = midpoint + (span / 2)
self.check_pcs_uutc_ss()
self.check_data_map_uutc_ss()
self.set_plot_data()
def scale_plot_data(self, up=True, scale=2):
for pc in self.get_plot_containers():
if up:
pc.scale_factor = pc.scale_factor * scale
else:
pc.scale_factor = pc.scale_factor / scale
self.update_signals()
return
def autoscale_plot_data(self, pc):
amp_span = np.abs(np.nanmax(pc.data)
- np.nanmin(pc.data))
row_span = 1 / self.visible_channels.get_row_count()
pc.scale_factor = row_span / (amp_span * pc.ufact)
def highlight_signal(self, pc):
# Determine the signal rectangle
w_step = 1/self.visible_channels.get_col_count()
h_step = 1/self.visible_channels.get_row_count()
# Pixel rel size to plot in inner rectangle
pix_w = 1/self.signal_view.width
pix_h = 1/self.signal_view.height
rel_sig_w = pc.plot_position[0] * w_step
rel_sig_h = pc.plot_position[1] * h_step
left = rel_sig_w + (3*pix_w)
right = rel_sig_w + w_step - (2*pix_w)
bottom = rel_sig_h + (2*pix_h)
top = rel_sig_h + h_step - (3*pix_h)
pos = np.array([[left, bottom, 10],
[right, bottom, 10],
[right, top, 10],
[left, top, 10]])
self.highlight_rec.set_data(vertices=pos,
color=np.concatenate([pc.line_color[:-1],
[0.2]]))
return
def apply_settings(self):
# update CONF attributes from init
self.canvas.bgcolor = CONF.get(self.CONF_SECTION,'bgcolor')
self.color_palette = CONF.get(self.CONF_SECTION, 'color_palette')
def __init__(self, viewer):
super(LassoSelectionMode, self).__init__(viewer)
self.line = Line(color='purple',
width=2,
method='agg',
parent=self._vispy_widget.canvas.scene)
print(open_mesh_boundary)
print('Here the mesh has a single boundary')
# WARNING : BrainObj should be added first before
visb_sc = splt.visbrain_plot(mesh=open_mesh, caption='open mesh')
# create points with vispy
for bound in open_mesh_boundary:
points = open_mesh.vertices[bound]
s_rad = SourceObj('rad',
points,
color='red',
symbol='square',
radius_min=10)
visb_sc.add_to_subplot(s_rad)
lines = Line(pos=open_mesh.vertices[bound], width=10, color='b')
# wrap the vispy object using visbrain
l_obj = VispyObj('line', lines)
visb_sc.add_to_subplot(l_obj)
visb_sc.preview()
# # here is how to get the vertices that define the boundary of
# # a texture on a mesh
mesh = sio.load_mesh('data/example_mesh.gii')
tex_parcel = sio.load_texture('data/example_texture_parcel.gii')
# bound_verts = stop.texture_boundary_vertices(tex_parcel.darray[0], 20, mesh.vertex_neighbors)
boundaries = list()
for val in np.unique(tex_parcel.darray[0]):
boundary = stop.texture_boundary(mesh, tex_parcel.darray[0], val)
boundaries.append(boundary)