Source code for glue.viewers.scatter.state

# -*- coding: utf-8 -*-

import numpy as np

from glue.core import BaseData, Subset

from glue.config import colormaps
from glue.viewers.matplotlib.state import (MatplotlibDataViewerState,
                                           DeferredDrawCallbackProperty as DDCProperty,
                                           DeferredDrawSelectionCallbackProperty as DDSCProperty)
from glue.core.state_objects import StateAttributeLimitsHelper
from echo import keep_in_sync, delay_callback
from glue.core.data_combo_helper import ComponentIDComboHelper, ComboHelper
from glue.core.exceptions import IncompatibleAttribute

from matplotlib.projections import get_projection_names

__all__ = ['ScatterViewerState', 'ScatterLayerState']

[docs]class ScatterViewerState(MatplotlibDataViewerState): """ A state class that includes all the attributes for a scatter viewer. """ x_att = DDSCProperty(docstring='The attribute to show on the x-axis', default_index=0) y_att = DDSCProperty(docstring='The attribute to show on the y-axis', default_index=1) dpi = DDCProperty(72, docstring='The resolution (in dots per inch) of density maps, if present') plot_mode = DDSCProperty(docstring="Whether to plot the data in cartesian, polar or another projection") angle_unit = DDSCProperty(docstring="Whether to use radians or degrees for any angular coordinates") def __init__(self, **kwargs): super(ScatterViewerState, self).__init__() self.limits_cache = {} self.x_lim_helper = StateAttributeLimitsHelper(self, attribute='x_att', lower='x_min', upper='x_max', log='x_log', margin=0.04, limits_cache=self.limits_cache) self.y_lim_helper = StateAttributeLimitsHelper(self, attribute='y_att', lower='y_min', upper='y_max', log='y_log', margin=0.04, limits_cache=self.limits_cache) self.add_callback('layers', self._layers_changed) self.x_att_helper = ComponentIDComboHelper(self, 'x_att', pixel_coord=True, world_coord=True) self.y_att_helper = ComponentIDComboHelper(self, 'y_att', pixel_coord=True, world_coord=True) self.plot_mode_helper = ComboHelper(self, 'plot_mode') self.plot_mode_helper.choices = [proj for proj in get_projection_names() if proj not in ['3d', 'scatter_density']] self.plot_mode_helper.selection = 'rectilinear' self.angle_unit_helper = ComboHelper(self, 'angle_unit') self.angle_unit_helper.choices = ['radians', 'degrees'] self.angle_unit_helper.selection = 'radians' self.update_from_dict(kwargs) self.add_callback('x_log', self._reset_x_limits) self.add_callback('y_log', self._reset_y_limits) if self.using_polar: self.full_circle() def _reset_x_limits(self, *args): if self.x_att is None: return self.x_lim_helper.percentile = 100 self.x_lim_helper.update_values(force=True) def _reset_y_limits(self, *args): if self.y_att is None: return self.y_lim_helper.percentile = 100 self.y_lim_helper.update_values(force=True)
[docs] def reset_limits(self): if not self.using_polar: self._reset_x_limits() self._reset_y_limits()
[docs] def flip_x(self): """ Flip the x_min/x_max limits. """ self.x_lim_helper.flip_limits()
[docs] def flip_y(self): """ Flip the y_min/y_max limits. """ self.y_lim_helper.flip_limits()
[docs] @property def using_rectilinear(self): return self.plot_mode == 'rectilinear'
[docs] @property def using_polar(self): return self.plot_mode == 'polar'
[docs] @property def using_full_sphere(self): return self.plot_mode in ['aitoff', 'hammer', 'mollweide', 'lambert']
[docs] @property def using_degrees(self): return (self.using_polar or self.using_full_sphere) and self.angle_unit == 'degrees'
[docs] @property def using_radians(self): return not self.using_rectilinear and self.angle_unit == 'radians'
[docs] def full_circle(self): if not self.using_polar: return self.x_min = 0 self.x_max = 2 * np.pi
[docs] @property def x_categories(self): return self._categories(self.x_att)
[docs] @property def y_categories(self): return self._categories(self.y_att)
def _categories(self, cid): categories = [] for layer_state in self.layers: if isinstance(layer_state.layer, BaseData): layer = layer_state.layer else: layer = try: if == 'categorical': categories.append( except IncompatibleAttribute: pass if len(categories) == 0: return None else: return np.unique(np.hstack(categories))
[docs] @property def x_kinds(self): return self._component_kinds(self.x_att)
[docs] @property def y_kinds(self): return self._component_kinds(self.y_att)
def _component_kinds(self, cid): # Construct list of component kinds over all layers kinds = set() for layer_state in self.layers: if isinstance(layer_state.layer, BaseData): layer = layer_state.layer else: layer = try: kinds.add( except IncompatibleAttribute: pass return kinds def _layers_changed(self, *args): layers_data = self.layers_data layers_data_cache = getattr(self, '_layers_data_cache', []) if layers_data == layers_data_cache: return self.x_att_helper.set_multiple_data(self.layers_data) self.y_att_helper.set_multiple_data(self.layers_data) self._layers_data_cache = layers_data
def display_func_slow(x): if x == 'Linear': return 'Linear (WARNING: may be slow due to data size)' else: return x
[docs]class ScatterLayerState(MatplotlibLayerState): """ A state class that includes all the attributes for layers in a scatter plot. """ # Color cmap_mode = DDSCProperty(docstring="Whether to use color to encode an attribute") cmap_att = DDSCProperty(docstring="The attribute to use for the color") cmap_vmin = DDCProperty(docstring="The lower level for the colormap") cmap_vmax = DDCProperty(docstring="The upper level for the colormap") cmap = DDCProperty(docstring="The colormap to use (when in colormap mode)") # Points points_mode = DDSCProperty(docstring='Whether to use markers or a density map') # Markers markers_visible = DDCProperty(True, docstring="Whether to show markers") size = DDCProperty(docstring="The size of the markers") size_mode = DDSCProperty(docstring="Whether to use size to encode an attribute") size_att = DDSCProperty(docstring="The attribute to use for the size") size_vmin = DDCProperty(docstring="The lower level for the size mapping") size_vmax = DDCProperty(docstring="The upper level for the size mapping") size_scaling = DDCProperty(1, docstring="Relative scaling of the size") fill = DDCProperty(True, docstring="Whether to fill the markers") # Density map density_map = DDCProperty(False, docstring="Whether to show the points as a density map") stretch = DDSCProperty(default='log', docstring='The stretch used to render the layer, ' 'which should be one of ``linear``, ' '``sqrt``, ``log``, or ``arcsinh``') density_contrast = DDCProperty(1, docstring="The dynamic range of the density map") # Note that we keep the dpi in the viewer state since we want it to always # be in sync between layers. # Line line_visible = DDCProperty(False, docstring="Whether to show a line connecting all positions") linewidth = DDCProperty(1, docstring="The line width") linestyle = DDSCProperty(docstring="The line style") # Errorbars xerr_visible = DDCProperty(False, docstring="Whether to show x error bars") yerr_visible = DDCProperty(False, docstring="Whether to show y error bars") xerr_att = DDSCProperty(docstring="The attribute to use for the x error bars") yerr_att = DDSCProperty(docstring="The attribute to use for the y error bars") # Vectors vector_visible = DDCProperty(False, docstring="Whether to show vector plot") vx_att = DDSCProperty(docstring="The attribute to use for the x vector arrow") vy_att = DDSCProperty(docstring="The attribute to use for the y vector arrow") vector_arrowhead = DDCProperty(False, docstring="Whether to show vector arrow") vector_mode = DDSCProperty(default_index=0, docstring="Whether to plot the vectors in cartesian or polar mode") vector_origin = DDSCProperty(default_index=1, docstring="Whether to place the vector so that the origin is at the tail, middle, or tip") vector_scaling = DDCProperty(1, docstring="The relative scaling of the arrow length") def __init__(self, viewer_state=None, layer=None, **kwargs): super(ScatterLayerState, self).__init__(viewer_state=viewer_state, layer=layer) self.limits_cache = {} self.cmap_lim_helper = StateAttributeLimitsHelper(self, attribute='cmap_att', lower='cmap_vmin', upper='cmap_vmax', limits_cache=self.limits_cache) self.size_lim_helper = StateAttributeLimitsHelper(self, attribute='size_att', lower='size_vmin', upper='size_vmax', limits_cache=self.limits_cache) self.cmap_att_helper = ComponentIDComboHelper(self, 'cmap_att', numeric=True, datetime=False, categorical=False) self.size_att_helper = ComponentIDComboHelper(self, 'size_att', numeric=True, datetime=False, categorical=False) self.xerr_att_helper = ComponentIDComboHelper(self, 'xerr_att', numeric=True, datetime=False, categorical=False) self.yerr_att_helper = ComponentIDComboHelper(self, 'yerr_att', numeric=True, datetime=False, categorical=False) self.vx_att_helper = ComponentIDComboHelper(self, 'vx_att', numeric=True, datetime=False, categorical=False) self.vy_att_helper = ComponentIDComboHelper(self, 'vy_att', numeric=True, datetime=False, categorical=False) self.points_mode_helper = ComboHelper(self, 'points_mode') points_mode_display = {'auto': 'Density map or markers (auto)', 'markers': 'Markers', 'density': 'Density map'} ScatterLayerState.points_mode.set_choices(self, ['auto', 'markers', 'density']) ScatterLayerState.points_mode.set_display_func(self, points_mode_display.get) self.add_callback('points_mode', self._update_density_map_mode) self.add_callback('density_map', self._on_density_map_change, priority=10000) ScatterLayerState.cmap_mode.set_choices(self, ['Fixed', 'Linear']) ScatterLayerState.size_mode.set_choices(self, ['Fixed', 'Linear']) linestyle_display = {'solid': '–––––––', 'dashed': '– – – – –', 'dotted': '· · · · · · · ·', 'dashdot': '– · – · – ·'} ScatterLayerState.linestyle.set_choices(self, ['solid', 'dashed', 'dotted', 'dashdot']) ScatterLayerState.linestyle.set_display_func(self, linestyle_display.get) ScatterLayerState.vector_mode.set_choices(self, ['Cartesian', 'Polar']) vector_origin_display = {'tail': 'Tail of vector', 'middle': 'Middle of vector', 'tip': 'Tip of vector'} ScatterLayerState.vector_origin.set_choices(self, ['tail', 'middle', 'tip']) ScatterLayerState.vector_origin.set_display_func(self, vector_origin_display.get) stretch_display = {'linear': 'Linear', 'sqrt': 'Square Root', 'arcsinh': 'Arcsinh', 'log': 'Logarithmic'} ScatterLayerState.stretch.set_choices(self, ['linear', 'sqrt', 'arcsinh', 'log']) ScatterLayerState.stretch.set_display_func(self, stretch_display.get) if self.viewer_state is not None: self.viewer_state.add_callback('x_att', self._on_xy_change, priority=10000) self.viewer_state.add_callback('y_att', self._on_xy_change, priority=10000) if hasattr(self.viewer_state, 'plot_mode'): self.viewer_state.add_callback('plot_mode', self._update_points_mode, priority=10000) self._on_xy_change() self._update_points_mode() self.add_callback('layer', self._on_layer_change) if layer is not None: self._on_layer_change() self.cmap = colormaps.members[0][1] self.size = self._sync_size = keep_in_sync(self, 'size',, 'markersize') self.update_from_dict(kwargs) def _update_points_mode(self, *args): if getattr(self.viewer_state, 'using_polar', False) or getattr(self.viewer_state, 'using_full_sphere', False): self.points_mode_helper.choices = ['markers'] = 'markers' else: self.points_mode_helper.choices = ['auto', 'markers', 'density'] def _on_xy_change(self, *event): if self.viewer_state.x_att is None or self.viewer_state.y_att is None: return if isinstance(self.layer, BaseData): layer = self.layer else: layer = try: x_datetime = layer.get_kind(self.viewer_state.x_att) == 'datetime' except IncompatibleAttribute: x_datetime = False try: y_datetime = layer.get_kind(self.viewer_state.y_att) == 'datetime' except IncompatibleAttribute: y_datetime = False with delay_callback(self, 'xerr_visible', 'yerr_visible', 'vector_visible'): if x_datetime: self.xerr_visible = False if y_datetime: self.yerr_visible = False if x_datetime or y_datetime: self.vector_visible = False def _on_layer_change(self, layer=None): with delay_callback(self, 'cmap_vmin', 'cmap_vmax', 'size_vmin', 'size_vmax', 'density_map'): self._update_density_map_mode() if self.layer is None: self.cmap_att_helper.set_multiple_data([]) self.size_att_helper.set_multiple_data([]) else: self.cmap_att_helper.set_multiple_data([self.layer]) self.size_att_helper.set_multiple_data([self.layer]) if self.layer is None: self.xerr_att_helper.set_multiple_data([]) self.yerr_att_helper.set_multiple_data([]) else: self.xerr_att_helper.set_multiple_data([self.layer]) self.yerr_att_helper.set_multiple_data([self.layer]) if self.layer is None: self.vx_att_helper.set_multiple_data([]) self.vy_att_helper.set_multiple_data([]) else: self.vx_att_helper.set_multiple_data([self.layer]) self.vy_att_helper.set_multiple_data([self.layer]) def _update_density_map_mode(self, *args): if self.points_mode == 'auto': if self.layer.size > 100000: self.density_map = True else: self.density_map = False elif self.points_mode == 'density': self.density_map = True else: self.density_map = False def _on_density_map_change(self, *args): # If the density map mode is used, we should disable the lines/errors/vectors if self.density_map: with delay_callback(self, 'line_visible', 'xerr_visible', 'yerr_visible', 'vector_visible'): if self.line_visible: self.line_visible = False if self.xerr_visible: self.xerr_visible = False if self.yerr_visible: self.yerr_visible = False if self.vector_visible: self.vector_visible = False
[docs] def flip_cmap(self): """ Flip the cmap_vmin/cmap_vmax limits. """ self.cmap_lim_helper.flip_limits()
[docs] def flip_size(self): """ Flip the size_vmin/size_vmax limits. """ self.size_lim_helper.flip_limits()
[docs] @property def cmap_name(self): return colormaps.name_from_cmap(self.cmap)
[docs] def compute_density_map(self, bins=None, range=None): if not self.markers_visible or not self.density_map: return np.zeros(bins) if isinstance(self.layer, Subset): data = subset_state = self.layer.subset_state else: data = self.layer subset_state = None count = data.compute_histogram([self.viewer_state.y_att, self.viewer_state.x_att], subset_state=subset_state, bins=bins, log=(self.viewer_state.y_log, self.viewer_state.x_log), range=range) if self.cmap_mode == 'Fixed': return count else: total = data.compute_histogram([self.viewer_state.y_att, self.viewer_state.x_att], subset_state=subset_state, bins=bins, weights=self.cmap_att, log=(self.viewer_state.y_log, self.viewer_state.x_log), range=range) return total / count
@classmethod def __setgluestate__(cls, rec, context): # Patch for glue files produced with glue v0.11 if 'style' in rec['values']: style = context.object(rec['values'].pop('style')) if style == 'Scatter': rec['values']['markers_visible'] = True rec['values']['line_visible'] = False elif style == 'Line': rec['values']['markers_visible'] = False rec['values']['line_visible'] = True return super(ScatterLayerState, cls).__setgluestate__(rec, context)