def _df_changed(self, old, new):
''' Handles how updates occur when df changes. Evaluates if columns or columnlabels
have been changed. Provides entry condition as well.
Note: New automatically sets self.df, so when I refer to new, I am actually
referring to self.df. Using "new" instead of self.df is just for readability'''
### Initialize plot first time df is passed into the class. Boolean listeners
### for df behave oddly, so uses self.plotdata for entry condition.
if not self.plotdata:
self.plotdata = PandasPlotData(df=new)
self._dffull = new
### Try to infer plot title from df name ###
try:
self.title = new.name #Turn into a "transfer" method where I pass list of attributes i want look for
except AttributeError:
pass
### Draw barebones of plot
self._plot_default()
self._update_lines()
### Decide to update columns or completely redraw df.
else:
changed = self.plotdata.set_df(new)
### If 'event changed', don't bother updaing lines
if not changed:
self._update_lines()
def _df_changed(self, old, new):
''' Handles how updates occur when df changes. Evaluates if columns or columnlabels
have been changed. Provides entry condition as well.
Note: New automatically sets self.df, so when I refer to new, I am actually
referring to self.df. Using "new" instead of self.df is just for readability'''
### Initialize plot first time df is passed into the class. Boolean listeners
### for df behave oddly, so uses self.plotdata for entry condition.
if not self.plotdata:
self.plotdata=PandasPlotData(df=new)
self._dffull=new
### Try to infer plot title from df name ###
try:
self.title=new.name #Turn into a "transfer" method where I pass list of attributes i want look for
except AttributeError:
pass
### Draw barebones of plot
self._plot_default()
self._update_lines()
### Decide to update columns or completely redraw df.
else:
changed=self.plotdata.set_df(new)
### If 'event changed', don't bother updaing lines
if not changed:
self._update_lines()
class PandasPlot(HasTraits):
'''Plot to interface a pandas df'''
### For testing
rnd_cols=Bool(False)
samecols=Bool(False)
transpose=Bool(False)
### Traits of the actual plot and plotdatasource handler ###
_dffull=Instance(DataFrame) #Any?
plot=Instance(Plot)
plotdata=Instance(PandasPlotData) #Data stored at any given time in the plot
df=Instance(DataFrame)
### Axis traits
idxname=Str('Index')
colname=Str('Column')
idxorient=DefaultEnum('top', 'bottom', 'left', 'right', default='bottom')
colorient=DefaultEnum('top', 'bottom', 'left', 'right', default='top')
selection_axis=DefaultEnum('index','value', default='index') #Which axis does selection tool sample
### Aesthetic traits
title=Str('Title') #Defaults from df name
linecolor=Str('yellow')
pointcolor=Str('red')
linestyle=Enum('line', 'scatter', 'both')
markersize=Range(low=1, high=5)
sampling=Range(low=0.0, high=100., value=100.0)
_spacing=Property(depends_on='sampling')
### _Event Handlers
def _rnd_cols_changed(self):
x=np.linspace(0, 100, 100) #Index generator
y=np.linspace(0, 100, 100) #Column generator
scale=randint(1,1000)
self.df=DataFrame((np.random.rand(len(x),len(y))), columns=(scale/2.0)*y, index=scale*x)
def _samecols_changed(self):
self.df=DataFrame((np.random.rand(100,100)))
def _transpose_changed(self):
self.df=self.df.transpose()
### Axis Aesthetics
@on_trait_change('idxname','idxcolumn', 'idxorient', 'colorient')
def _axis_changed(self, trait, new):
'''Change plot axis name or orientation'''
setattr(self, trait, new)
self._update_axis()
self.plot.request_redraw() #Necessary but how do I only call axis redraw?
def _title_changed(self, old, new):
if old != new:
self.plot.title=new
self.plot.request_redraw() #HOW TO PREVENT COLLISION WITH TOP OVERLAY
def _selection_axis_changed(self, old, new):
if old != new:
self._update_lines()
def _df_changed(self, old, new):
''' Handles how updates occur when df changes. Evaluates if columns or columnlabels
have been changed. Provides entry condition as well.
Note: New automatically sets self.df, so when I refer to new, I am actually
referring to self.df. Using "new" instead of self.df is just for readability'''
### Initialize plot first time df is passed into the class. Boolean listeners
### for df behave oddly, so uses self.plotdata for entry condition.
if not self.plotdata:
self.plotdata=PandasPlotData(df=new)
self._dffull=new
### Try to infer plot title from df name ###
try:
self.title=new.name #Turn into a "transfer" method where I pass list of attributes i want look for
except AttributeError:
pass
### Draw barebones of plot
self._plot_default()
self._update_lines()
### Decide to update columns or completely redraw df.
else:
changed=self.plotdata.set_df(new)
### If 'event changed', don't bother updaing lines
if not changed:
self._update_lines()
### Plot spacing
def _get__spacing(self):
'''Integer spacing given % sampling'''
samp=self.sampling/100.0
colsize=df.shape[0]
return colsize - int(round( colsize*samp, 0))
def _sampling_changed(self, old, new):
if old != new:
self._update_samples()
def _update_samples(self):
'''Updates the list of line plots shown or hidden based on plot sampling.'''
if self._spacing==0:
return
### Hide all plots
to_hide=self.plot.plots.keys()
self.plot.hideplot(*to_hide)
### Show only plots in samples
to_show=self.plot.plots.keys()[::self._spacing]
self.plot.showplot(*to_show)
self.plot.request_redraw()
def _update_rangeselect(self):
''' Overwrites range selection tool in current plot.'''
#### Remove current overlay
self.plot.overlays=[obj for obj in self.plot.overlays if not isinstance(obj, RangeSelectionOverlay)]
mycomp=self.plot.plots.itervalues().next()[0] #Quick wayt to get first value in dictionary
inds=range(len(self.df.index))
idx=ArrayDataSource(inds)
vals=ArrayDataSource(df.index.values)
index_range = DataRange1D(idx)
val_range=DataRange1D(vals)
imap=LinearMapper(range=index_range)#,stretch_data=self.index_mapper.stretch_data)
vmap=LinearMapper(range=val_range)
# mycomp.index_range.refresh()
mycomp.index_mapper=imap
mycomp.value_mapper=vmap
self.rangeselect=RangeSelection(mycomp, axis=self.selection_axis)
self.plot.active_tool = self.rangeselect
self.plot.overlays.append(RangeSelectionOverlay(component=mycomp))
self.rangeselect.on_trait_change(self.on_selection_changed, "selection")
def _update_lines(self):
''' Redraws lines, plots and reapplies line selection.'''
oldplots=self.plot.plots.keys()
newplots=[name for name in self.plotdata.list_data(as_strings=True)]
to_remove=[p for p in oldplots if p not in newplots]
to_add=[p for p in newplots if p not in oldplots]
if to_remove:
for p in to_remove:
self.plot.delplot(p)
if to_add:
for name in to_add:
self.plot.plot(('index', name), name=name, color=self.linecolor)
self._update_axis()
self._update_samples()
self._update_rangeselect()
def on_selection_changed(self, selection):
if selection != None:
self.rangeXMin, self.rangeXMax = selection
print selection
def _update_axis(self):
''' Forces a label axis onto the plot. '''
print 'updaing axis', self.idxname
indexlabels=[str(round(i,1)) for i in self.df.index]
columnlabels=[str(round(i,1)) for i in self.df.columns]
index_axis=LabelAxis(self.plot, orientation=self.idxorient,
positions=range(int(float(indexlabels[0])),
int(float(indexlabels[-1]))),
labels=indexlabels,#, resizable='hv',
title=self.idxname)
col_axis=LabelAxis(self.plot, orientation=self.colorient,
positions=range(int(float(columnlabels[0])),
int(float(columnlabels[-1]))),
labels=columnlabels,#, resizable='hv',
title=self.colname)
### Remove underlays
self.plot.underlays=[obj for obj in self.plot.underlays if not isinstance(obj, PlotAxis)]
self.plot.underlays=[obj for obj in self.plot.underlays if not isinstance(obj, LabelAxis)]
self.plot.underlays.append(index_axis)
self.plot.underlays.append(col_axis)
def _plot_default(self, toolbar=True, **pltkwds):
''' Draw bare plot, including main plotting area, toolbar, etc...
either at initialization or global redo'''
if toolbar:
self.plot=ToolbarPlot(self.plotdata, **pltkwds)
else:
self.plot=Plot(self.plotdata, **pltkwds)
self.plot.title = self.title
self.plot.padding = 50
self.plot.legend.visible=False
self.plot.tools.append(PanTool(self.plot))
zoom=BetterSelectingZoom(component=self.plot, tool_mode="box", always_on=False)
self.plot.overlays.append(zoom)
def _overwrite_plotdata(self):
'''When a new instance of PandasPlotData is created, this overwrites the
data source and updates the axis values.'''
self.plotdata=PandasPlotData(df=self.df)
self._plot_default() #CAN THIS JUST DRAW LINES
### Traits View
main_group=Group(
HGroup(Item('rnd_cols'), Item('samecols'), Item('transpose'),),
Item('plot', editor=ComponentEditor(), show_label=False),
Item('idxname'), Item('idxorient'), Item('selection_axis'), Item('title'),
Item('sampling'),
# Item('df_new'), Item('df_change'),
#Include('sample_group'),
#Include('axis_traits_group')
)
traits_view=View( Include('main_group') , height=600, width=800)
def _overwrite_plotdata(self):
'''When a new instance of PandasPlotData is created, this overwrites the
data source and updates the axis values.'''
self.plotdata=PandasPlotData(df=self.df)
self._plot_default() #CAN THIS JUST DRAW LINES
class PandasPlot(HasTraits):
'''Plot to interface a pandas df'''
### For testing
rnd_cols = Bool(False)
samecols = Bool(False)
transpose = Bool(False)
### Traits of the actual plot and plotdatasource handler ###
_dffull = Instance(DataFrame) #Any?
plot = Instance(Plot)
plotdata = Instance(
PandasPlotData) #Data stored at any given time in the plot
df = Instance(DataFrame)
### Axis traits
idxname = Str('Index')
colname = Str('Column')
idxorient = DefaultEnum('top', 'bottom', 'left', 'right', default='bottom')
colorient = DefaultEnum('top', 'bottom', 'left', 'right', default='top')
selection_axis = DefaultEnum(
'index', 'value',
default='index') #Which axis does selection tool sample
### Aesthetic traits
title = Str('Title') #Defaults from df name
linecolor = Str('yellow')
pointcolor = Str('red')
linestyle = Enum('line', 'scatter', 'both')
markersize = Range(low=1, high=5)
sampling = Range(low=0.0, high=100., value=100.0)
_spacing = Property(depends_on='sampling')
### _Event Handlers
def _rnd_cols_changed(self):
x = np.linspace(0, 100, 100) #Index generator
y = np.linspace(0, 100, 100) #Column generator
scale = randint(1, 1000)
self.df = DataFrame((np.random.rand(len(x), len(y))),
columns=(scale / 2.0) * y,
index=scale * x)
def _samecols_changed(self):
self.df = DataFrame((np.random.rand(100, 100)))
def _transpose_changed(self):
self.df = self.df.transpose()
### Axis Aesthetics
@on_trait_change('idxname', 'idxcolumn', 'idxorient', 'colorient')
def _axis_changed(self, trait, new):
'''Change plot axis name or orientation'''
setattr(self, trait, new)
self._update_axis()
self.plot.request_redraw(
) #Necessary but how do I only call axis redraw?
def _title_changed(self, old, new):
if old != new:
self.plot.title = new
self.plot.request_redraw(
) #HOW TO PREVENT COLLISION WITH TOP OVERLAY
def _selection_axis_changed(self, old, new):
if old != new:
self._update_lines()
def _df_changed(self, old, new):
''' Handles how updates occur when df changes. Evaluates if columns or columnlabels
have been changed. Provides entry condition as well.
Note: New automatically sets self.df, so when I refer to new, I am actually
referring to self.df. Using "new" instead of self.df is just for readability'''
### Initialize plot first time df is passed into the class. Boolean listeners
### for df behave oddly, so uses self.plotdata for entry condition.
if not self.plotdata:
self.plotdata = PandasPlotData(df=new)
self._dffull = new
### Try to infer plot title from df name ###
try:
self.title = new.name #Turn into a "transfer" method where I pass list of attributes i want look for
except AttributeError:
pass
### Draw barebones of plot
self._plot_default()
self._update_lines()
### Decide to update columns or completely redraw df.
else:
changed = self.plotdata.set_df(new)
### If 'event changed', don't bother updaing lines
if not changed:
self._update_lines()
### Plot spacing
def _get__spacing(self):
'''Integer spacing given % sampling'''
samp = self.sampling / 100.0
colsize = df.shape[0]
return colsize - int(round(colsize * samp, 0))
def _sampling_changed(self, old, new):
if old != new:
self._update_samples()
def _update_samples(self):
'''Updates the list of line plots shown or hidden based on plot sampling.'''
if self._spacing == 0:
return
### Hide all plots
to_hide = self.plot.plots.keys()
self.plot.hideplot(*to_hide)
### Show only plots in samples
to_show = self.plot.plots.keys()[::self._spacing]
self.plot.showplot(*to_show)
self.plot.request_redraw()
def _update_rangeselect(self):
''' Overwrites range selection tool in current plot.'''
#### Remove current overlay
self.plot.overlays = [
obj for obj in self.plot.overlays
if not isinstance(obj, RangeSelectionOverlay)
]
mycomp = self.plot.plots.itervalues().next()[
0] #Quick wayt to get first value in dictionary
inds = range(len(self.df.index))
idx = ArrayDataSource(inds)
vals = ArrayDataSource(df.index.values)
index_range = DataRange1D(idx)
val_range = DataRange1D(vals)
imap = LinearMapper(
range=index_range) #,stretch_data=self.index_mapper.stretch_data)
vmap = LinearMapper(range=val_range)
# mycomp.index_range.refresh()
mycomp.index_mapper = imap
mycomp.value_mapper = vmap
self.rangeselect = RangeSelection(mycomp, axis=self.selection_axis)
self.plot.active_tool = self.rangeselect
self.plot.overlays.append(RangeSelectionOverlay(component=mycomp))
self.rangeselect.on_trait_change(self.on_selection_changed,
"selection")
def _update_lines(self):
''' Redraws lines, plots and reapplies line selection.'''
oldplots = self.plot.plots.keys()
newplots = [name for name in self.plotdata.list_data(as_strings=True)]
to_remove = [p for p in oldplots if p not in newplots]
to_add = [p for p in newplots if p not in oldplots]
if to_remove:
for p in to_remove:
self.plot.delplot(p)
if to_add:
for name in to_add:
self.plot.plot(('index', name),
name=name,
color=self.linecolor)
self._update_axis()
self._update_samples()
self._update_rangeselect()
def on_selection_changed(self, selection):
if selection != None:
self.rangeXMin, self.rangeXMax = selection
print selection
def _update_axis(self):
''' Forces a label axis onto the plot. '''
print 'updaing axis', self.idxname
indexlabels = [str(round(i, 1)) for i in self.df.index]
columnlabels = [str(round(i, 1)) for i in self.df.columns]
index_axis = LabelAxis(
self.plot,
orientation=self.idxorient,
positions=range(int(float(indexlabels[0])),
int(float(indexlabels[-1]))),
labels=indexlabels, #, resizable='hv',
title=self.idxname)
col_axis = LabelAxis(
self.plot,
orientation=self.colorient,
positions=range(int(float(columnlabels[0])),
int(float(columnlabels[-1]))),
labels=columnlabels, #, resizable='hv',
title=self.colname)
### Remove underlays
self.plot.underlays = [
obj for obj in self.plot.underlays
if not isinstance(obj, PlotAxis)
]
self.plot.underlays = [
obj for obj in self.plot.underlays
if not isinstance(obj, LabelAxis)
]
self.plot.underlays.append(index_axis)
self.plot.underlays.append(col_axis)
def _plot_default(self, toolbar=True, **pltkwds):
''' Draw bare plot, including main plotting area, toolbar, etc...
either at initialization or global redo'''
if toolbar:
self.plot = ToolbarPlot(self.plotdata, **pltkwds)
else:
self.plot = Plot(self.plotdata, **pltkwds)
self.plot.title = self.title
self.plot.padding = 50
self.plot.legend.visible = False
self.plot.tools.append(PanTool(self.plot))
zoom = BetterSelectingZoom(component=self.plot,
tool_mode="box",
always_on=False)
self.plot.overlays.append(zoom)
def _overwrite_plotdata(self):
'''When a new instance of PandasPlotData is created, this overwrites the
data source and updates the axis values.'''
self.plotdata = PandasPlotData(df=self.df)
self._plot_default() #CAN THIS JUST DRAW LINES
### Traits View
main_group = Group(
HGroup(
Item('rnd_cols'),
Item('samecols'),
Item('transpose'),
),
Item('plot', editor=ComponentEditor(), show_label=False),
Item('idxname'),
Item('idxorient'),
Item('selection_axis'),
Item('title'),
Item('sampling'),
# Item('df_new'), Item('df_change'),
#Include('sample_group'),
#Include('axis_traits_group')
)
traits_view = View(Include('main_group'), height=600, width=800)
def _overwrite_plotdata(self):
'''When a new instance of PandasPlotData is created, this overwrites the
data source and updates the axis values.'''
self.plotdata = PandasPlotData(df=self.df)
self._plot_default() #CAN THIS JUST DRAW LINES