def display_plots(filebase, directory=None, width=700, height=500, **kwargs):
"""Display a series of plots controlled by sliders. The function relies on Python string format functionality to index through a series of plots."""
def show_figure(filebase, directory, **kwargs):
"""Helper function to load in the relevant plot for display."""
filename = filebase.format(**kwargs)
if directory is not None:
filename = directory + '/' + filename
display(HTML("<img src='{filename}'>".format(filename=filename)))
interact(show_figure, filebase=fixed(filebase), directory=fixed(directory), **kwargs)
def setFloat(self, label, default=0, min=-20, max=20, description='Set Float', format='text'):
""" Set float in a notebook pipeline (via interaction or with nbconvert) """
obj = self.load(label)
if obj == None:
obj=default
self.save(obj, label) # initialize with default
floatw = FloatSlider(value=obj, min=min, max=max, description=description)
hndl = interact(self.save, obj=floatw, label=fixed(label), format=fixed(format))
def setText(self, label, default='', description='Set Text', format='text'):
""" Set text in a notebook pipeline (via interaction or with nbconvert) """
obj = self.load(label)
if obj == None:
obj=default
self.save(obj, label) # initialize with default
textw = Text(value=obj, description=description)
hndl = interact(self.save, obj=textw, label=fixed(label), format=fixed(format))
def setDropdown(self, label, default=None, options=[], description='Set Dropdown', format='text'):
""" Set float in a notebook pipeline (via interaction or with nbconvert) """
obj = self.load(label)
if obj == None:
obj=default
self.save(obj, label) # initialize with default
dropdownw = Dropdown(value=obj, options=options, description=description)
hndl = interact(self.save, obj=dropdownw, label=fixed(label), format=fixed(format))
def layer_select(m, design):
layers_to_plot = m
features_to_plot = {'features':[],'layer':[]}
if len(layers_to_plot)>0:
features_to_plot['features'] = [x for x in design.features if design.features[x].layer in layers_to_plot]
features_to_plot['layers'] = [layers_to_plot.index(design.features[x].layer) for x in design.features if design.features[x].layer in layers_to_plot]
if len(features_to_plot['features'])>0:
#chk = [Checkbox(description=a) for a in features_to_plot]
#interact(updatePlot, **{c.description: c.value for c in chk})
sel_mult2 = SelectMultiple(options = features_to_plot['features'])
interactive_plot2 = interactive(updatePlot, n=sel_mult2,
design = fixed(design), all_features = fixed(features_to_plot))
display(interactive_plot2)
def im(self, z=None, zmin=None, zmax=None, cmap='jet'):
if z is None:
if self._interactive:
name_options = list()
for dim_names, var_names in self._inspector.dim_names_to_var_names.items():
no_zero_dim = all([self._inspector.dim_name_to_size[dim] > 0 for dim in dim_names])
if no_zero_dim and len(dim_names) == 2:
name_options.extend(var_names)
name_options = sorted(name_options)
# TODO (forman, 20160709): add sliders for zmin, zmax
interact(self._plot_im, z_name=name_options, zmin=fixed(zmax), zmax=fixed(zmax), cmap=fixed(cmap))
else:
raise ValueError('name must be given')
else:
self._plot_im(z_name=z, zmin=zmin, zmax=zmax, cmap=cmap)
def myinteract(img):
min_ = round(np.nanmin(img), 4)
max_ = round(np.nanmax(img), 4)
p30, p70 = np.percentile(img[~np.isnan(img)], (30, 70))
delta = round((p30-min_)/50, 5)
interact(myimshow, img=fixed(img), vmin=(min_, p30, delta),
vmax=(p70, max_, delta), i=(0, len(interpolators)-1))
def plot_expm(r:(0,4,0.1)=1.1, n0=fixed(100)):
nt = [r**t * n0 for t in range(10)]
plt.figure(figsize=[9,4])
plt.plot(nt, lw=2)
plt.xlabel('generation number')
plt.ylabel('population size')
plt.ylim(0, max(nt)*1.1)
plt.show()
def test_fixed():
c = interactive(f, a=widgets.fixed(5), b='text')
nt.assert_equal(len(c.children), 2)
w = c.children[0]
check_widget(w,
cls=widgets.Text,
value='text',
description='b',
)
def plot_mc_sol(s1:(0,1,0.1)=1,p:(0,1,0.1)=0.9,q:(0,1,0.1)=0.9, tmax=fixed(20)):
s1,s2 = list(zip(*[mc_sol(s1,p,q,t) for t in range(tmax+1)]))
plt.figure(figsize=[9,4])
plt.plot(s1, label='S_1, off')
plt.plot(s2, label='S_2, on')
plt.xlabel('time')
plt.ylabel('proportion of channels')
plt.ylim(-0.01,1.01)
plt.legend()
plt.show()
def display_widget(f, title_to_filename, select_widget, checkbox_list):
'''
this function creates the interactive window for the user to explore the data
'''
i = widgets.interactive(f,
movie_title = select_widget,
title_to_filename = fixed(title_to_filename),
anger=checkbox_list[0],
anticipation=checkbox_list[1],
disgust=checkbox_list[2],
fear=checkbox_list[3],
joy=checkbox_list[4],
negative=checkbox_list[5],
positive=checkbox_list[6],
sadness=checkbox_list[7],
surprise=checkbox_list[8],
trust=checkbox_list[9],
visulalize_emotions=fixed(True),
hierarchy_emotions=widgets.ToggleButton(value=False,
description='See emotion importance for\
the whole script'),
hierarchy_emotions_subplot=widgets.ToggleButton(value=False,
description='See emotion importance for\
the beginning, middle and end of the script'))
hbox1 = widgets.HBox([i.children[0]]) #select movie to explore
hbox2 = widgets.HBox(i.children[6:8]) #positive and negative sentiment analysis
hbox3 = widgets.HBox(i.children[1:6]+i.children[8:11]) #emotions
hbox4 = widgets.HBox(i.children[11:12])
hbox5 = widgets.HBox(i.children[12:])
display( hbox1 )
display( hbox2 )
display( hbox3 )
display( hbox4 )
display( hbox5 )
def plot_mc_sol2(p:(0,1,0.1)=0.9,q:(0,1,0.1)=0.9, \
tmax=fixed(20), log_n:(0,10,1)=0):
s1,s2 = list(zip(*[mc_sol(1,p,q,t) for t in range(tmax+1)]))
n = int(np.exp(log_n))
sim = [list(mc(p,q,tmax)) for _ in range(n)]
sim_av = [np.mean(s) for s in list(zip(*sim))]
print("n = %d" % n)
plt.figure(figsize=[9,4])
plt.plot(s2)
plt.plot(sim_av)
plt.xlabel('time')
plt.ylabel('proportion of channels on')
plt.ylim(-0.01,1.01)
plt.show()
def plot_riemann_SW(h_l,h_r,u_l,u_r,g=1.,force_waves=None,extra_lines=None, tracer=False):
stripes = not tracer
plot_function_stripes, plot_function_xt_phase = \
make_plot_functions(h_l,h_r,u_l,u_r,g,
force_waves,extra_lines,stripes=stripes)
interact(plot_function_stripes,
t=widgets.FloatSlider(value=0.,min=0,max=.9), fig=fixed(0))
if tracer:
interact(plot_function_xt_phase,
plot_1_chars=Checkbox(description='1-characteristics',
value=False),
plot_2_chars=Checkbox(description='2-characteristics'),
plot_tracer_chars=Checkbox(description='Tracer characteristics'))
else:
interact(plot_function_xt_phase,
plot_1_chars=Checkbox(description='1-characteristics',
value=False),
plot_2_chars=Checkbox(description='2-characteristics'))
def plot_1(thedata):
widgets.interact(_plot_1,
thedata = widgets.fixed(thedata),
Country1 = widgets.Dropdown(
description = 'OECD Country (No data for % AVG Wage for Turkey)',
options = thedata['Country'].unique().tolist(),
value = 'Australia',
disabled = False),
Country2 = widgets.Dropdown(
description = 'OECD Country (No data for % AVG Wage for Turkey)',
options = thedata['Country'].unique().tolist(),
value = 'Australia',
disabled = False),
Variable1 = widgets.Dropdown(
description = 'Variable1',
options = ['Total Unemployment (%)','Inflation Rate (%)','Average Wage Growth (%)','GDP Growth (%)'],
value = 'Total Unemployment (%)'),
)
def show_results(results, header, prefix):
"""!@brief Display results
A function to display all the accepted candidates using a dropdown list.
@param results (np.ndarray) numpy array of results sorted by total energy
@param header (str) A comma separated string of the output properties of the conformer
@param prefix (dict) A dictionary of the prefix of the run and the associated helical configuration
@returns None
@sa run
@sa single_result
"""
# Set a list of conformer names and their indices in the results
options = [(str(int(conformer[0])) + '_' + str(int(conformer[1])) + '.pdb', i) for i, conformer in enumerate(results)]
# Show a dropdown widget of all the accepted conformers
dropdown = widgets.Dropdown(value=options[0][1], options=options, style={'description_width': 'initial'}, description='Conformer')
w = widgets.interactive(single_result, result=dropdown, header=widgets.fixed(header), results=widgets.fixed(results), prefix=widgets.fixed(prefix))
display(w)
def confined_aquifer(analysis=False):
approx = Checkbox(value=True, description='approx.')
semilog = Checkbox(value=False, description='SemiLog')
Q = FloatSlider(value=1000,
description=r'$Q$ [m$^3$/day]',
min=500,
max=1500,
step=500,
continuous_update=False)
t = FloatLogSlider(value=1.0,
description=r'$t$ [day]',
base=10,
min=-1,
max=2,
step=0.2,
continuous_update=False)
r = FloatSlider(value=200,
description=r'$r$ [m]',
min=100,
max=500,
step=100,
continuous_update=False)
T = FloatSlider(value=300,
description=r'$T$ [m$^2$/day]',
min=100,
max=500,
step=100,
continuous_update=False)
io = interactive_output(
show_theis, {
'Q': Q,
't': t,
'r': r,
'T': T,
'approx': approx,
'semilog': semilog,
'analysis': fixed(analysis)
})
return VBox([HBox([Q, t, approx]), HBox([T, r, semilog]), io])
def bases(param):
"""!@brief Bases widget for use in Jupyter notebook
This function displays information on the available nucleobases that are defined in the program.
The bases are defined in "data/bases_library.yaml"
@returns None
@sa display_options_widgets
"""
display(widgets.HTML(value='<H3>Bases</H3>'))
display(widgets.HTML(value=('These bases are already defined:' +
'<br> Adenine (A), Guanine (G), Cytosine (C), Uracil (U), Thymine (T), ' +
'Cyanuric Acid (I), and aminopyrimidine (E)')))
num_defined_bases = len([k for k in param if 'Base' in k]) - 1 # Subtract the item from the main options
w = widgets.interactive(add_base, param=widgets.fixed(param), number_of_bases=widgets.BoundedIntText(value=num_defined_bases, min=0,
description="Number of additional bases", style={'description_width': 'initial'}))
help_box = widgets.Button(description='?', tooltip=('The number of additional bases to define'), layout=widgets.Layout(width='4%'))
display(widgets.HBox([help_box, w]))
def interactive_indifference_curves(preferences="cobb_douglas", **kwargs):
interactive_utility_settings(preferences, kwargs)
widgets.interact(
_interactive_indifference_curves,
alpha=widgets.FloatSlider(
description="$\\alpha$",
min=kwargs["alpha_min"],
max=kwargs["alpha_max"],
step=kwargs["alpha_step"],
value=kwargs["alpha"],
),
beta=widgets.FloatSlider(
description="$\\beta$",
min=kwargs["beta_min"],
max=kwargs["beta_max"],
step=kwargs["beta_step"],
value=kwargs["beta"],
),
par=widgets.fixed(kwargs),
)
def create_wordcloud(frame = fixed(text) , community = [0, 1, 2, 3], maximum = [50,100,300] , atitle = fixed('')):
if community == 0:
mk = imageio.imread('./data/club.jpg')
if community == 1:
mk = imageio.imread('./data/diamond.png')
if community == 2:
mk = imageio.imread('./data/spade.jpg')
if community == 3:
mk = imageio.imread('./data/heart.jpg')
wordcloud = WordCloud(max_font_size = 70,
max_words = maximum,
background_color = 'white',
collocations = False,
mask = mk).generate(text[community])
image_colors = ImageColorGenerator(mk)
plt.style.use('seaborn')
plt.figure(figsize=[15, 15])
plt.title('community '+str(community), fontsize = 20)
plt.imshow(wordcloud.recolor(color_func = image_colors), interpolation = 'bilinear')
plt.axis('off')
def nifti_plotter(self,
plotting_func=None,
colormap=None,
figsize=(15, 5),
**kwargs):
"""
Plot volumetric data.
Args
----
plotting_func : function
A plotting function for .nii files, most likely
mask_background : bool
Whether the background should be masked (set to NA).
This parameter only works in conjunction with the default
plotting function (`plotting_func=None`). It finds clusters
of values that round to zero and somewhere touch the edges
of the image. These are set to NA. If you think you are
missing data in your image, set this False.
colormap : str | list
The matplotlib colormap that should be applied to the data.
By default, the widget will allow you to pick from all that
are available, but you can pass a string to fix the
colormap or a list of strings to offer the user a few
options.
figsize : tup
The figure height and width for matplotlib, in inches.
If you are providing a custom plot function, any kwargs you provide
to nifti_plotter will be passed to that function.
"""
kwargs['colormap'] = get_cmap_dropdown(colormap)
kwargs['figsize'] = fixed(figsize)
if plotting_func is None:
self._default_plotter(**kwargs)
else:
self._custom_plotter(plotting_func, **kwargs)
def interact():
from ipywidgets import interactive, fixed, widgets, interact_manual
from IPython.display import display
from ipywidgets import AppLayout, Button, Layout
target_values_options = list(df["{}"].unique())
features_options = list(df.columns.values)
features_options.remove("{}")
w1 = widgets.SelectMultiple(
options=target_values_options,
value=target_values_options,
rows=len(target_values_options),
description="Target Values (column: {})",
style={{
"description_width": "initial"
}},
disabled=False,
)
w2 = widgets.SelectMultiple(
options=features_options,
value=features_options[0:min(len(features_options), 2)],
rows=len(features_options),
description="Features",
disabled=False,
)
ui = widgets.HBox([w1, w2])
out = widgets.interactive_output(
pairwise_scatter_plots,
{{
"df": fixed(df),
"target_values": w1,
"features": w2
}},
)
display(ui, out)
def display_lineplot_widget(cls, df, include_all=False):
options = df.StationCode.unique().tolist()
stations_dropdown = cls.create_stations_dropdown(options, include_all)
validation_dropdown = cls.create_validation_dropdown()
options = df.select_dtypes("float").columns.tolist()
param_dropdown = cls.create_param_dropdown(options)
options = df.DateTimeStamp.dt.year.unique().tolist()
year_dropdown = cls.create_year_dropdown(options)
def on_dropdown_update(df, station_code, validation_code, param_code, year_code):
df = filter_by_station_code(df, station_code)
df = filter_by_year_code(df, year_code)
df = filter_by_validation_code(df, validation_code)
if not df.empty:
msg = f"Generating lineplot for {station_code}: {validation_code}"
display(widgets.HTML(msg))
lineplot_options = dict(
x="DateTimeStamp",
y=param_code,
hue="StationCode",
ci=None
)
cls.create_lineplot(df, **lineplot_options)
else:
msg = "No data was found with the selected filters."
msg += "Unable to generate report."
display(widgets.HTML(msg))
widgets.interact(
on_dropdown_update,
df=widgets.fixed(df),
station_code=stations_dropdown,
validation_code=validation_dropdown,
param_code=param_dropdown,
year_code=year_dropdown
)
def observation_error():
#out = Output()
seed = 13
rolldice = Button(description='ROLL THE DICE',
tooltip='randomise the random number generator')
Nsldr = IntSlider(value=5,
description='$N_{obs}$',
min=2,
max=10,
step=1,
continuous_update=False)
trueModel = Checkbox(value=False, description='True Process')
bestModel = Checkbox(value=False, description='Best (LSQ) Model')
varsldr = FloatLogSlider(value=0.1,
base=10,
description='$\sigma_i^2$',
min=-2,
max=0,
step=1,
continuous_update=False)
sdf = fixed(seed)
np.random.seed(13)
def on_button_clicked(b):
sdf.value = int(time.time())
rolldice.on_click(on_button_clicked)
io = interactive_output(
plot_observations, {
'N_obs': Nsldr,
'trueModel': trueModel,
'bestModel': bestModel,
'var': varsldr,
'seed': sdf
})
return VBox([HBox([Nsldr, bestModel, trueModel, rolldice, varsldr]), io])
def plot_measures(data_path):
dd_style = {'description_width': 'initial'}
w = ipywidgets.Dropdown(options=[
'dikeraising_evr_smooth', 'dikeraising_lrg_smooth',
'groynelowering_evr_smooth', 'groynelowering_lrg_smooth',
'lowering_evr_natural', 'lowering_evr_smooth', 'lowering_lrg_natural',
'lowering_lrg_smooth', 'minemblowering_evr_smooth',
'minemblowering_lrg_smooth', 'sidechannel_evr_natural',
'sidechannel_evr_smooth', 'sidechannel_lrg_natural',
'sidechannel_lrg_smooth', 'smoothing_evr_natural',
'smoothing_evr_smooth', 'smoothing_lrg_natural', 'smoothing_lrg_smooth'
],
value='lowering_evr_smooth',
description='Select measure:',
style=dd_style)
res = interactive(_plot_datasets,
data_path=fixed(data_path),
measure_dir=w)
display(res)
def im_line(self, z=None, zmin=None, zmax=None, xind=None, yind=None, cmap='jet'):
if self._interactive:
has_xind = xind is not None
has_yind = yind is not None
if not z or not has_xind or not has_yind:
if xind is None:
widget_xind = widgets.IntSlider(min=0, max=10, step=1, description='X:')
else:
widget_xind = fixed(xind)
if yind is None:
widget_yind = widgets.IntSlider(min=0, max=10, step=1, description='Y:')
else:
widget_yind = fixed(yind)
if not z:
valid_var_names = list()
for dim_names, var_names in self._inspector.dim_names_to_var_names.items():
no_zero_dim = all([self._inspector.dim_name_to_size[dim] > 0 for dim in dim_names])
if no_zero_dim and len(dim_names) == 2:
valid_var_names.extend(var_names)
valid_var_names = sorted(valid_var_names)
value = z if z and z in valid_var_names else valid_var_names[0]
widget_var = widgets.Dropdown(options=valid_var_names, value=value, description='Var:')
# noinspection PyUnusedLocal
def on_widget_var_change(change):
variable = self._inspector.dataset[widget_var.value]
if xind is None:
widget_xind.max = variable.shape[1] - 1
if yind is None:
widget_yind.max = variable.shape[0] - 1
widget_var.observe(on_widget_var_change, names='value')
else:
widget_var = fixed(z)
# TODO (forman, 20160709): add sliders for zmin, zmax
interact(self._plot_im_line, z_name=widget_var,
xind=widget_xind, yind=widget_yind,
zmin=fixed(zmax), zmax=fixed(zmax), cmap=fixed(cmap))
else:
if not z:
raise ValueError('z_name must be given')
self._plot_im_line(z, zmin=zmin, zmax=zmin, xind=xind, yind=yind, cmap=cmap)
def visualise(self, sample):
encoder_ids = sample['encoder_input_ids'].flatten()
decoder_ids = sample['decoder_input_ids']
all_ids = encoder_ids.tolist() + decoder_ids.tolist()
encoder_tokens = self.tokenizer.batch_decode(encoder_ids.view(-1, 1))
decoder_tokens = self.tokenizer.batch_decode(decoder_ids.view(-1, 1))
all_tokens = encoder_tokens + decoder_tokens
all_token_and_ids = [
f"{idx}->{tok}" for idx, tok in zip(all_ids, all_tokens)
]
ipyw.interact(self.visualise_for_token,
sample=ipyw.fixed(sample),
target_id=ipyw.SelectionSlider(
options=all_token_and_ids,
value=all_token_and_ids[0],
description='Selected token:',
disabled=False,
continuous_update=False,
orientation='horizontal',
readout=True))
def prediction(var):
mtrue, ctrue = [2, 3]
cmin, c0, cmax = [2.55, 3.05, 3.55]
mmin, m0, mmax = [1.3, 2.1, 2.9]
cmin, c0, cmax = [1.55, 3.05, 4.55]
mmin, m0, mmax = [0.3, 2.1, 4.9]
p = Posterior(cmin=cmin,
cmax=cmax,
Nc=101,
mmin=mmin,
mmax=mmax,
Nm=101,
ctrue=ctrue,
mtrue=mtrue,
var=var)
zoom = Checkbox(value=False, description='zoom')
Nsamples = FloatLogSlider(value=16,
base=4,
description='samples',
min=0,
max=5,
step=1,
continuous_update=False)
xf = FloatSlider(value=3,
description=r'$x_f$',
min=2,
max=5,
step=0.5,
continuous_update=False)
io = interactive_output(plot_predictions, {
'zoom': zoom,
'N': Nsamples,
'xf': xf,
'p': fixed(p)
})
return VBox([HBox([zoom, Nsamples, xf]), io])
def paginate_df(df, nb_items=10):
def show_df(df, page):
display(df.head(page * nb_items).tail(nb_items))
def show_next(change):
pagination_slider.value += change
def get_pagination_buttons():
next_button = widgets.Button(layout=widgets.Layout(width='30px'),
icon='chevron-right')
next_button.on_click(lambda _: show_next(1))
prev_button = widgets.Button(layout=widgets.Layout(width='30px'),
icon='chevron-left')
prev_button.on_click(lambda _: show_next(-1))
return [prev_button, next_button]
nb_rows = widgets.Label(value='{} rows'.format(len(df)))
if len(df) > nb_items:
nb_pages = int(np.ceil(len(df) / nb_items))
pagination_slider = widgets.IntSlider(value=1,
min=1,
max=nb_pages,
layout=Layout(width='60%'))
pagination_slider_label = widgets.Label(
value='of {} pages with '.format(nb_pages))
pagination_controls = widgets.HBox([
pagination_slider, pagination_slider_label, nb_rows,
*get_pagination_buttons()
])
paginated_table = interactive_output(
show_df, dict(df=fixed(df), page=pagination_slider))
display(pagination_controls, paginated_table)
else:
display(nb_rows, df)
def manual_ps(data, notebook=False):
"""
Manual Phase correction using matplotlib
A matplotlib widget is used to manually correct the phase of a Fourier
transformed dataset. If the dataset has more than 1 dimensions, the first
trace will be picked up for phase correction. Clicking the 'Set Phase'
button will print the current linear phase parameters to the console.
A ipywidget is provided for use with Jupyter Notebook to avoid changing
backends. This can be accessed with notebook=True option in this function
.. note:: Needs matplotlib with an interactive backend.
Parameters
----------
data : ndarray
Array of NMR data.
notebook : Bool
True for plotting interactively in Jupyter Notebook
Uses ipywidgets instead of matplotlib widgets
Returns
-------
p0, p1 : float
Linear phase correction parameters. Zero and first order phase
corrections in degrees calculated from pc0, pc1 and pivot displayed
in the interactive window.
Examples
--------
>>> import nmrglue as ng
>>> p0, p1 = ng.process.proc_autophase.manual_ps(data)
>>> # do manual phase correction and close window
>>> phased_data = ng.proc_base.ps(data, p0=p0, p1=p1)
In [1] # if you are using the Jupyter Notebook
In [2] ng.process.proc_autophase.manual_ps(data)
Out [2] # do manual phase correction. p0 and p1 values will be updated
# continuously as you do so and are printed below the plot
In [3] phased_data = ng.proc_base.ps(data, p0=p0, p1=p1)
"""
if len(data.shape) == 2:
data = data[0, ...]
elif len(data.shape) == 3:
data = data[0, 0, ...]
elif len(data.shape) == 4:
data = data[0, 0, 0, ...]
if notebook:
from ipywidgets import interact, fixed
import matplotlib.pyplot as plt
def phasecorr(dataset, phcorr0, phcorr1, pivot):
fig, ax = plt.subplots(figsize=(10, 7))
phaseddata = dataset * np.exp(
1j * (phcorr0 + phcorr1 * (
np.arange(-pivot, -pivot+dataset.size)/dataset.size)))
ax.plot(np.real(phaseddata))
ax.set(ylim=(np.min(np.real(data))*2, np.max(np.real(data))*2))
ax.axvline(pivot, color='r', alpha=0.5)
plt.show()
p0 = np.round(
(phcorr0 - phcorr1 * pivot/dataset.size) * 360 / 2 / np.pi, 3)
p1 = np.round(phcorr1*360/2/np.pi, 3)
print('p0 =', p0, 'p1 =', p1)
interact(
phasecorr,
dataset=fixed(data),
phcorr0=(-np.pi, np.pi, 0.01),
phcorr1=(-10*np.pi, 10*np.pi, 0.01),
pivot=(0, data.size, 1))
else:
from matplotlib.widgets import Slider, Button
import matplotlib.pyplot as plt
plt.subplots_adjust(left=0.25, bottom=0.35)
interactive, = plt.plot(data.real, lw=1, color='black')
axcolor = 'white'
axpc0 = plt.axes([0.25, 0.10, 0.65, 0.03], facecolor=axcolor)
axpc1 = plt.axes([0.25, 0.15, 0.65, 0.03], facecolor=axcolor)
axpiv = plt.axes([0.25, 0.20, 0.65, 0.03], facecolor=axcolor)
axpst = plt.axes([0.25, 0.25, 0.15, 0.04], facecolor=axcolor)
spc0 = Slider(axpc0, 'p0', -360, 360, valinit=0)
spc1 = Slider(axpc1, 'p1', -360, 360, valinit=0)
spiv = Slider(axpiv, 'pivot', 0, data.size, valinit=0)
axps = Button(axpst, 'Set Phase', color=axcolor)
def update(val):
pc0 = spc0.val * np.pi / 180
pc1 = spc1.val * np.pi / 180
pivot = spiv.val
interactive.set_ydata((data * np.exp(
1.0j * (pc0 + (pc1 * np.arange(-pivot, -pivot + data.size) /
data.size))).astype(data.dtype)).real)
plt.draw()
def setphase(val):
p0 = spc0.val-spc1.val*spiv.val/data.size
p1 = spc1.val
print(p0, p1)
spc0.on_changed(update)
spc1.on_changed(update)
spiv.on_changed(update)
axps.on_clicked(setphase)
plt.show(block=True)
p0 = spc0.val-spc1.val*spiv.val/data.size
p1 = spc1.val
return p0, p1
def dynamic_lisa_composite_explore(rose, gdf, pattern='',
p=0.05, figsize=(13, 10)):
"""
Interactive exploration of dynamic LISA values
for different dates in a dataframe.
Note: only possible in jupyter notebooks
Parameters
----------
rose : giddy.directional.Rose instance
A ``Rose`` object, which contains (among other attributes)
weights to calculate `esda.moran.Moran_local` values
gdf : geopandas dataframe instance
The Dataframe containing information and polygons to plot.
pattern : str, optional
Option to extract all columns ending with a specific pattern.
Only extracted columns will be used for comparison.
p : float, optional
The p-value threshold for significance. Default =0.05
figsize: tuple, optional
W, h of figure. Default =(13,10)
Returns
-------
None
Examples
--------
**Note**: this function creates Jupyter notebook widgets, so is meant only
to run in a notebook.
>>> import geopandas as gpd
>>> import pandas as pd
>>> from pysal.lib.weights.contiguity import Queen
>>> from pysal.lib import examples
>>> import numpy as np
>>> import matplotlib.pyplot as plt
If you want to see figures embedded inline in a Jupyter notebook,
add a line ``%matplotlib inline`` at the top of your notebook.
>>> from pysal.explore.giddy.directional import Rose
>>> from pysal.viz.splot.giddy import dynamic_lisa_composite_explore
get csv and shp files
>>> shp_link = examples.get_path('us48.shp')
>>> df = gpd.read_file(shp_link)
>>> income_table = pd.read_csv(examples.get_path("usjoin.csv"))
calculate relative values
>>> for year in range(1969, 2010):
... income_table[str(year) + '_rel'] = (
... income_table[str(year)] / income_table[str(year)].mean())
merge to one gdf
>>> gdf = df.merge(income_table,left_on='STATE_NAME',right_on='Name')
retrieve spatial weights and data for two points in time
>>> w = Queen.from_dataframe(gdf)
>>> w.transform = 'r'
>>> y1 = gdf['1969_rel'].values
>>> y2 = gdf['2000_rel'].values
calculate rose Object
>>> Y = np.array([y1, y2]).T
>>> rose = Rose(Y, w, k=5)
plot
>>> fig = dynamic_lisa_composite_explore(rose, gdf, pattern='rel')
>>> # plt.show()
"""
coldict = {col: col for col in gdf.columns if
col.endswith(pattern)}
interact(_dynamic_lisa_widget_update,
start_time=coldict, end_time=coldict, rose=fixed(rose),
gdf=fixed(gdf), p=fixed(p), figsize=fixed(figsize))
def waveform_line(self, ind=None, ref_ind=None):
"""
Draw waveform 2D line plot.
:param ind: Time index
:param ref_ind: Reference time index
"""
if ind is None and self._interactive:
interact(self._plot_waveform_line, ind=(0, self._inspector.num_times - 1), ref_ind=fixed(ref_ind))
else:
self._plot_waveform_line(ind=ind if ind else 0, ref_ind=ref_ind)
def _default_plotter(self, mask_background=False, **kwargs):
"""Plot three orthogonal views.
This is called by nifti_plotter, you shouldn't call it directly.
"""
plt.gcf().clear()
plt.ioff() # disable interactive mode
data_array = self.data.get_data()
if not ((data_array.ndim == 3) or (data_array.ndim == 4)):
raise ValueError('Input image should be 3D or 4D')
# mask the background
if mask_background:
# TODO: add the ability to pass 'mne' to use a default brain mask
# TODO: split this out into a different function
if data_array.ndim == 3:
labels, n_labels = scipy.ndimage.measurements.label(
(np.round(data_array) == 0))
else: # 4D
labels, n_labels = scipy.ndimage.measurements.label(
(np.round(data_array).max(axis=3) == 0))
mask_labels = [
lab for lab in range(1, n_labels + 1)
if (np.any(labels[[0, -1], :, :] == lab)
| np.any(labels[:, [0, -1], :] == lab)
| np.any(labels[:, :, [0, -1]] == lab))
]
if data_array.ndim == 3:
data_array = np.ma.masked_where(np.isin(labels, mask_labels),
data_array)
else:
data_array = np.ma.masked_where(
np.broadcast_to(
np.isin(labels, mask_labels)[:, :, :, np.newaxis],
data_array.shape), data_array)
# init sliders for the various dimensions
for dim, label in enumerate(['x', 'y', 'z']):
if label not in kwargs.keys():
kwargs[label] = IntSlider(value=(data_array.shape[dim] - 1) /
2,
min=0,
max=data_array.shape[dim] - 1,
continuous_update=False)
if (data_array.ndim == 3) or (data_array.shape[3] == 1):
kwargs['t'] = fixed(None) # time is fixed
else:
kwargs['t'] = IntSlider(value=0,
min=0,
max=data_array.shape[3] - 1,
continuous_update=False)
widgets.interact(self._plot_slices, data=fixed(data_array), **kwargs)
plt.close() # clear plot
plt.ion() # return to interactive state
merged.set_index(['continent','year'],inplace=True)
merged = merged.join(merged_grouped_last)
merged.reset_index(inplace=True)
merged['g_total'] = merged['last']/merged['first']*100
# Dropdown widget on pop and gdp_cap development:
def plot(dataframe, continent):
I = dataframe['continent'] == continent
ax_gdp = dataframe.loc[I,:].plot(x='year', y = 'gdp', style = '-o', legend = 'False')
ax_pop = dataframe.loc[I,:].plot(x='year', y = 'pop', style = '-o', legend = 'False')
widgets.interact(plot,
dataframe = widgets.fixed(merged),
continent = widgets.Dropdown(description='continent', options=merged.continent.unique(), value='Europe & Central Asia')
);
# Creating 5-year growth rates in GDP per cap in the continents:
merged_5 = merged[merged['year'].isin(['1970','1975','1980','1985','1990','1995','2000','2005','2010','2015'])]
merged_5.set_index(['continent','year'],inplace=True)
merged_5_grouped = merged_5.groupby('continent')
merged_5_change = merged_5_grouped.gdp_cap.pct_change()
merged_5_change.name = 'growth_5'
merged_5 = merged_5.join(merged_5_change)
merged_5.reset_index(inplace=True)
# Bar plot with dropdown widget:
def cr72_manipulate(pA=None,
r20a=None,
r20b=None,
kex=None,
dw_h=None,
dw_n=None,
num=None):
"""
My attempt at using interact in a function.
INPUTS:
r20a, r20b: intrinsic r2 rates for states a and b, respectively, usually
the same.
pA: population of the major state.
kex: k1 + k2, the sum of the forward and reverse rates of exchange.
dw_h, dw_n: differences in chem. shift between the two states in ppm for
1H and 15N, respectively
TO DO:
Add in the ability to import parameter values from the `relax` analysis
for a particular residue and set the values and ranges in the interact
call automatically.
Make another function to plot RD data for a given residue and then a third
function that calls both the data and simulation functions. I MAY HAVE THE
FIRST FUNCTION HERE IN MY RD PLOTTING FROM BEFORE.
Add in functionality to calculate whether a residue is in slow, fast, or
intermediate exchange. That could be a function itself.
"""
import sys
sys.path.append('/usr/local/relax-4.0.2/lib')
sys.path.append('/usr/local/relax-4.0.2')
import relax
import nmr
from dispersion.cr72 import r2eff_CR72
from ipywidgets import interact, interactive, fixed
import ipywidgets as widgets
import matplotlib.pyplot as plt
import numpy as np
# The number of steps for each slider
if num == None:
num = 15
# Choosing the major population slider
if pA == None:
lower, upper = .5, 1.
step, val = (upper - lower) / num, 0.9
pA_slider = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
else:
lower, upper = .5, 1.
step, val = (upper - lower) / num, pA
pA_slider = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
# Setting the intrinsic r2 value, held constant
if r20a == None or r20b == None:
r2int = fixed(20)
elif r20b == None:
r2int = fixed(r20a)
else:
r2int = fixed(r20b)
# Setting kex
if kex == None:
lower, upper, step, val = 100., 600., (600. - 100.) / num, 500.
kex_slider = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
else:
lower, upper = .1 * kex, 1.5 * kex
step, val = (upper - lower) / num, kex
kex_slider = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
# Setting dw_h & dw_n
if dw_h == None or dw_n == None:
lower, upper = .1, 4.
step, val = (upper - lower) / num, 1.
dw_n_slid = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
lower, upper = .01, 1
step, val = (upper - lower) / num, .1
dw_h_slid = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
else:
lower, upper = dw_h * .1, dw_h * 1.5
step, val = (upper - lower) / num, dw_h
dw_h_slid = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
lower, upper = dw_n * .1, dw_n * 1.5
step, val = (upper - lower) / num, dw_n
dw_n_slid = widgets.FloatSlider(min=lower,
max=upper,
step=step,
value=val)
interact(cr72_simulate,
pA=pA_slider,
r20a=r2int,
r20b=r2int,
kex=kex_slider,
dw_h=dw_h_slid,
dw_n=dw_n_slid)
return
def qonduit_visualization_metrics_plot_histogram(data):
return interactive(lambda data: display(plot_state_histogram(data)),
data=fixed(data))
min=0.01,
max=0.1,
step=0.01,
value=.05,
continuous_update=True,
readout_format='.3f',
description='Init. capital ratio',
style = {'description_width': 'initial'})
# %% tags=[]
# Make the widget
interact(plot1,
Epsilon = Epsilon_widget1,
DiscFac = DiscFac_widget1,
PopGrowth = PopGrowth_widget1,
YearsPerGeneration = fixed(years_per_gen),
Initialk = Initialk_widget1);
# %% [markdown]
# ### Gross and Net Per Capita Output as a Function of k
# %%
# Define a function that plots something given some inputs
def plot2(Epsilon, PopGrowth, YearsPerGeneration):
'''Inputs:
Epsilon: Elasticity of output with respect to capital/labour ratio
DiscFac: One period discount factor
YearPerGeneration: No. of years per generation
PopGrowth: Gross growth rate of population in one period'''
def interactive_rad(temp_pd, stations):
"""Interactive plotting of elevation-azimuth plots for a set of satellites and a set
of stations
Parameters
----------
temp_pd : Pandas dataframe
Output of aiub.import_RM_file function
stations: pandas dataframe
station dataframe as output of aiub.import_stations()
Returns
-------
"""
all_dates = temp_pd.time_stamp.unique(
) #temp_pd.datetime.apply(lambda x: x.timestamp()).unique()
#all_dates_read = temp_pd.datetime.unique()
time_list = list(
map(lambda x: datetime.datetime.fromtimestamp(x),
temp_pd.time_stamp.unique()))
all_dates_read = [
str(tl.year) + '-' + str(tl.month) + '-' + str(tl.day) + ' ' +
str(tl.hour) + ':' + str(tl.minute) for tl in time_list
]
colors = ['red', 'blue', 'green', 'cyan', 'orange']
items = [
widgets.ColorPicker(description=temp_pd.satellite.unique()[i],
value=colors[i])
for i in range(len(temp_pd.satellite.unique()))
]
if len(items) < 5:
items = items + [
widgets.ColorPicker(description='None', value=colors[i])
for i in range(5 - len(items))
]
colwidget = widgets.VBox(items)
#min_date_widget = widgets.IntSlider(min=0, max=len(all_dates)-1, step=1, value = 0, continuous_update = False)
#max_date_widget = widgets.IntSlider(min=0, max=len(all_dates)-1, step=1, value = 0, continuous_update = False)
style = {'description_width': '20%', 'readout_width': '20%'}
min_date_widget = widgets.SelectionSlider(options=all_dates_read,
style=style,
description='Min Time',
layout={'width': '400px'})
max_date_widget = widgets.SelectionSlider(options=all_dates_read,
style=style,
description='Max Time',
layout={'width': '400px'})
sat_type = widgets.SelectMultiple(options=temp_pd.satellite.unique(),
description='Select satellite types',
value=[temp_pd.satellite.unique()[0]],
disabled=False)
station_widget = widgets.Dropdown(options=stations.statname.unique(),
value=stations.iloc[0].statname,
description='Select observation station',
style={'description_width': '50%'},
layout={'width': '400px'})
d = {
'data': widgets.fixed(temp_pd),
'all_dates': widgets.fixed(all_dates),
'all_dates_read': widgets.fixed(all_dates_read),
'sat_type': sat_type,
'date_min': min_date_widget,
'date_max': max_date_widget,
'stations': widgets.fixed(stations),
'station_sel': station_widget
}
d2 = {'col' + str(ind): value for (ind, value) in enumerate(items)}
d.update(d2)
title_w = widgets.HTML(
"<span style='float:left;font-size:2em;font-weight:bold'>Radial plot of satellite observation</span>"
)
time_title_w = widgets.HTML(
"<span style='float:left;font-size:1em;font-weight:bold'>Pick time range for observations</span>"
)
w1 = widgets.HBox([sat_type, colwidget])
w2 = widgets.HBox([station_widget])
w3 = widgets.HBox([min_date_widget, max_date_widget])
ui = widgets.VBox([title_w, w1, w2, time_title_w, w3],
layout=Layout(align_items='stretch'))
out = widgets.interactive_output(plot_radial, d)
display(ui, out)
#Interact functionality with GUIs
#Works only on jupiter notebook
from ipywidgets import interact, interactive, fixed
import ipywidgets as widgets
def func(x):
return x
interact(func, x='Hello')
@interact(x=True, y=fixed(1.0))
def g(x, y):
return (x, y)
interact(func, x=widgets.IntSlider(min=-100, max=100, step=1, value=0))
interact(func, x=(-10, 10, .1))
@interact(x=(0.0, 20.0, 0.5))
def h(x=5.0):
return x
interact(func, x=['Hello', 'option 2', 'option 3'])
interact(func, x={'one': 10, 'two:': 20})
def _default_plotter(self, **kwargs):
"""
Basic plot function to be used if no custom function is specified.
This is called by plot, you shouldn't call it directly.
"""
self.lengths = np.array([length(x) for x in self.lines2use_])
if not ("grayscale" in kwargs and kwargs["grayscale"]):
self.colors = np.array([color(x) for x in self.lines2use_])
else:
self.colors = np.zeros((len(self.lines2use_), 3), dtype=np.float16)
self.colors[:] = [0.5, 0.5, 0.5]
self.state = {"threshold": 0, "indices": []}
width = 600
height = 600
perc = 80
if "width" in kwargs:
width = kwargs["width"]
if "height" in kwargs:
height = kwargs["height"]
if "percentile" in kwargs:
perc = kwargs["percentile"]
ipv.clear()
fig = ipv.figure(width=width, height=height)
self.state["fig"] = fig
with fig.hold_sync():
x, y, z, indices, colors, self.line_pointers = self._create_mesh()
limits = np.array(
[
min([x.min(), y.min(), z.min()]),
max([x.max(), y.max(), z.max()]),
]
)
mesh = ipv.Mesh(x=x, y=y, z=z, lines=indices, color=colors)
fig.meshes = [mesh]
if "style" not in kwargs:
fig.style = {
"axes": {
"color": "black",
"label": {"color": "black"},
"ticklabel": {"color": "black"},
"visible": False,
},
"background-color": "white",
"box": {"visible": False},
}
else:
fig.style = kwargs["style"]
ipv.pylab._grow_limits(limits, limits, limits)
fig.camera_fov = 1
ipv.show()
interact(
self._plot_lines,
state=fixed(self.state),
threshold=widgets.FloatSlider(
value=np.percentile(self.lengths, perc),
min=self.lengths.min() - 1,
max=self.lengths.max() - 1,
continuous_update=False,
),
)
def display_prediction(basis, num_basis=4, wlim=(-1.,1.), fig=None, ax=None, xlim=None, ylim=None, num_points=1000, offset=0.0, **kwargs):
"""Interactive widget for displaying a prediction function based on summing separate basis functions.
:param basis: a function handle that calls the basis functions.
:type basis: function handle.
:param xlim: limits of the x axis to use.
:param ylim: limits of the y axis to use.
:param wlim: limits for the basis function weights."""
import numpy as np
import pylab as plt
if fig is not None:
if ax is None:
ax = fig.gca()
if xlim is None:
if ax is not None:
xlim = ax.get_xlim()
else:
xlim = (-2., 2.)
if ylim is None:
if ax is not None:
ylim = ax.get_ylim()
else:
ylim = (-1., 1.)
# initialise X and set up W arguments.
x = np.zeros((num_points, 1))
x[:, 0] = np.linspace(xlim[0], xlim[1], num_points)
param_args = {}
for i in range(num_basis):
lim = list(wlim)
if i ==0:
lim[0] += offset
lim[1] += offset
param_args['w_' + str(i)] = lim
# helper function for making basis prediction.
def predict_basis(w, basis, x, num_basis, **kwargs):
Phi = basis(x, num_basis, **kwargs)
f = np.dot(Phi, w)
return f, Phi
if type(basis) is dict:
use_basis = basis[list(basis.keys())[0]]
else:
use_basis = basis
f, Phi = predict_basis(np.zeros((num_basis, 1)),
use_basis, x, num_basis,
**kwargs)
if fig is None:
fig, ax=plt.subplots(figsize=(12,4))
ax.set_ylim(ylim)
ax.set_xlim(xlim)
predline = ax.plot(x, f, linewidth=2)[0]
basislines = []
for i in range(num_basis):
basislines.append(ax.plot(x, Phi[:, i], 'r')[0])
ax.set_ylim(ylim)
ax.set_xlim(xlim)
def generate_function(basis, num_basis, predline, basislines, basis_args, display_basis, offset, **kwargs):
w = np.zeros((num_basis, 1))
for i in range(num_basis):
w[i] = kwargs['w_'+ str(i)]
f, Phi = predict_basis(w, basis, x, num_basis, **basis_args)
predline.set_xdata(x[:, 0])
predline.set_ydata(f)
for i in range(num_basis):
basislines[i].set_xdata(x[:, 0])
basislines[i].set_ydata(Phi[:, i])
if display_basis:
for i in range(num_basis):
basislines[i].set_alpha(1) # make visible
else:
for i in range(num_basis):
basislines[i].set_alpha(0)
display(fig)
if type(basis) is not dict:
basis = fixed(basis)
plt.close(fig)
interact(generate_function,
basis=basis,
num_basis=fixed(num_basis),
predline=fixed(predline),
basislines=fixed(basislines),
basis_args=fixed(kwargs),
offset = fixed(offset),
display_basis = False,
**param_args)
def single_run(data_dir="", id=""):
if not os.path.exists(data_dir):
raise FileNotFoundError("No directory found named: "+ data_dir)
run_summary = RunSummary(data_dir=data_dir, id=id)
sedov_solution = SedovSolution(E_0,
run_summary.overview.background_density,
run_summary.overview.metallicity)
#### PASS TO PLOTTER ####
num_checkpoints = len(run_summary.filenames)
# log_R_max = round(np.log10(run_summary.df["Radius"].max()), 2)
# log_R_min = max(log_R_max-4,
# round(np.log10(run_summary.df["Radius"].min()), 2)+1)
R_min = run_summary.df["Radius"].min()
R_min = 0
R_max = run_summary.df["Radius"].max()
if type(single_run.previous_widget) is widgets.Box:
single_run.previous_widget.close()
w = interactive(plotter,
run_summary = fixed(run_summary),
sedov_solution = fixed(sedov_solution),
xlim = FloatRangeSlider(min=R_min,
max=R_max,
step=0.01 * (R_max-R_min),
value=(R_min, R_max),
),
checkpoint_index = IntSlider(min=0,
max=num_checkpoints-1,
step=0,
value=num_checkpoints-1),
y_axis_variable = Dropdown(options=cols_plot,
value="Density"),
x_axis_variable = RadioButtons(options=["Radius",
"M_int",
"zones"]),
label = fixed("numeric"),
density_in_mH_cm3 = fixed(False),
verbose = fixed(True),
)
w1 = widgets.Button(description=u"\u23EA")
def show_first(b):
checkpoint_slider = w.children[5]
checkpoint_slider.value = checkpoint_slider.min
w1.on_click(show_first)
wl = widgets.Button(description=u"\u276E")
def show_prev(b):
checkpoint_slider = w.children[5]
if checkpoint_slider.value > checkpoint_slider.min:
checkpoint_slider.value -= 1
wl.on_click(show_prev)
w2 = widgets.Button(description=u"\u25BA")
# w2.stop = False
def play(b):
checkpoint_slider = w.children[5]
for i in range(checkpoint_slider.value+1, checkpoint_slider.max+1):
plt.gcf()
checkpoint_slider.value=i
# plt.show()
# time.sleep(.1)
# if b.stop:
# break
w2.on_click(play)
# wp = widgets.Button(description=u"p")
# def pause(b):
# w2.stop=True
# wp.on_click(pause)
wr = widgets.Button(description=u"\u276F")
def show_next(b):
checkpoint_slider = w.children[5]
if checkpoint_slider.value < checkpoint_slider.max:
checkpoint_slider.value += 1
wr.on_click(show_next)
w3 = widgets.Button(description=u"\u23E9")
def show_last(b):
checkpoint_slider = w.children[5]
checkpoint_slider.value = checkpoint_slider.max
w3.on_click(show_last)
w_buttons = widgets.HBox(children=[w1, wl, w2, wr, w3])
w.children = tuple([_w for _w in w.children ] + [w_buttons])
single_run.previous_widget = w
display(w)
# display(w_buttons)
return run_summary
def display_registration_results(fixed_image, moving_image, tx):
moving_resampled = sitk.Resample(moving_image, fixed_image, tx, sitk.sitkLinear, 0.0, moving_image.GetPixelIDValue())
interact(display_images_with_alpha, image_z=(0,fixed_image.GetSize()[2]-1), alpha=(0.0,1.0,0.05), image1 = fixed(fixed_image), image2=fixed(moving_resampled));
def display_gui(wti_index, print_args=False):
def lw(width='180px', left='0'):
return widgets.Layout(width=width)
def parties_options():
return ['ALL'] + sorted([ x for x in wti_index.get_countries_list() if x not in ['ALL', 'ALL OTHER']])
period_group_index_widget = widgets_config.period_group_widget(index_as_value=True)
topic_group_name_widget = widgets_config.topic_groups_widget(layout=lw())
parties_widget = widgets_config.parties_widget(options=parties_options(), value=['ALL'], rows=10)
recode_7corr_widget = widgets_config.toggle('Recode 7CULT', True, tooltip='Recode all treaties with cultural=yes as 7CORR', layout=lw(width='110px'))
use_lemma_widget = widgets_config.toggle('Use LEMMA', False, tooltip='Use WordNet lemma', layout=lw(width='110px'))
remove_stopwords_widget = widgets_config.toggle('Remove STOP', True, tooltip='Do not include stopwords', layout=lw(width='110px'))
compute_co_occurance_widget = widgets_config.toggle('Cooccurrence', True, tooltip='Compute Cooccurrence', layout=lw(width='110px'))
extra_groupbys_widget = widgets_config.dropdown('Groupbys', EXTRA_GROUPBY_OPTIONS, None, layout=lw())
min_word_size_widget = widgets_config.itext(0, 5, 2, description='Min word', layout=lw())
# plot_style_widget = widgets_config.plot_style_widget(layout=lw())
n_top_widget = widgets_config.slider('Top/grp #', 2, 100, 25, step=10, layout=lw(width='200px'))
n_min_count_widget = widgets_config.slider('Min count', 1, 10, 5, step=1, tooltip='Filter out words with count less than specified value', layout=lw(width='200px'))
treaty_filter_widget = widgets_config.dropdown('Filter', config.TREATY_FILTER_OPTIONS, 'is_cultural', layout=lw())
output_format_widget = widgets_config.dropdown('Output', OUTPUT_OPTIONS, 'plot_stacked_bar', layout=lw())
progress_widget = widgets_config.progress(0, 8, 1, 0, layout=lw(width='95%'))
def progress(x=None):
progress_widget.value = progress_widget.value + 1 if x is None else x
iw = widgets.interactive(
display_headnote_toplist,
period_group_index=period_group_index_widget,
topic_group_name=topic_group_name_widget,
recode_7corr=recode_7corr_widget,
treaty_filter=treaty_filter_widget,
parties=parties_widget,
extra_groupbys=extra_groupbys_widget,
n_min_count=n_min_count_widget,
n_top=n_top_widget,
min_word_size=min_word_size_widget,
use_lemma=use_lemma_widget,
compute_co_occurance=compute_co_occurance_widget,
remove_stopwords=remove_stopwords_widget,
output_format=output_format_widget,
progress=widgets.fixed(progress),
wti_index=widgets.fixed(wti_index),
print_args=widgets.fixed(print_args)
# plot_style=plot_style
)
boxes = widgets.HBox(
[
widgets.VBox([ period_group_index_widget, parties_widget ]),
widgets.VBox([ topic_group_name_widget, extra_groupbys_widget, n_top_widget, min_word_size_widget, n_min_count_widget]),
widgets.VBox([ recode_7corr_widget, use_lemma_widget, remove_stopwords_widget, compute_co_occurance_widget ]),
widgets.VBox([ treaty_filter_widget, output_format_widget, progress_widget]),
]
)
display(widgets.VBox([boxes, iw.children[-1]]))
iw.update()
disabled=False)
# Define a check box for whether to show the target annotation
show_targ_widget = widgets.Checkbox(value=True,
description='Show target $(m,c)$',
disabled=False)
# Make an interactive plot of the tractable buffer stock solution
# To make some of the widgets not appear, replace X_widget with fixed(desired_fixed_value) in the arguments below.
interact(
makeTBSplot,
DiscFac=DiscFac_widget,
CRRA=CRRA_widget,
# We can fix a parameter using the fixed() operator
Rfree=fixed(TBS_dictionary['Rfree']),
# Rfree = Rfree_widget, # This is the line which, when uncommented, would make Rfree a slider
PermGroFac=PermGroFac_widget,
UnempPrb=UnempPrb_widget,
mMin=mMin_widget,
mMax=mMax_widget,
cMin=cMin_widget,
cMax=cMax_widget,
show_targ=show_targ_widget,
plot_emp=plot_emp_widget,
plot_ret=plot_ret_widget,
plot_mSS=plot_mSS_widget,
)
# %%
def NM_fast(L_dev,
Y_dev,
k=2,
sig=0.01,
policy='new',
verbose=False,
return_more_info=False):
# create pd dataframe
Complete_dev = np.concatenate((np.array([Y_dev]).T, L_dev), axis=1)
df = pd.DataFrame(data=Complete_dev,
columns=["GT"] +
["LF_" + str(i) for i in range(L_dev.shape[1])])
no_other_LFs = L_dev.shape[1] - 2
def create_CT_tables(df, L_dev, k):
"""create all combinations of contingency table's of {LF_i, LF_j, [LF_all others]}"""
CT_list = []
for i in range(L_dev.shape[1]):
for j in [k for k in range(i, L_dev.shape[1]) if k != i]:
other_LFs_list = [
df['LF_' + str(m)] for m in range(L_dev.shape[1])
if (m != i and m != j)
]
CT = pd.crosstab([df['GT']] + other_LFs_list +
[df['LF_' + str(i)]],
df['LF_' + str(j)],
margins=False)
#k_list = [i-1 for i in range(k+1)]; GT_indices = [i for i in range(k)]
#CT = CT.reindex(index=list(itertools.product(GT_indices, *[tuple(k_list)] * (no_other_LFs+1))), columns=k_list, fill_value=0)
# prep to reindex only the LF column closest to values (this is new in the fast version)
indices_other_LFs = [
] # first get current indices of other LF columns
for itera in range(len(CT.index)):
indices_other_LFs.append(CT.index[itera][:-1])
indices_other_LFs = list(
set(indices_other_LFs)) # get unique values only
indices_closest_LF = [(i - 1, ) for i in range(k + 1)]
all_indices = [
ind1 + ind2 for ind1 in indices_other_LFs
for ind2 in indices_closest_LF
]
k_list = [i - 1 for i in range(k + 1)]
# reindex only the LF column closest to values
CT = CT.reindex(index=all_indices,
columns=k_list,
fill_value=0)
CT_list.append(CT)
return CT_list
def show_CT(q, CT_list):
"""function to return qth CT at index q-1 of CT_list"""
return CT_list[q - 1]
# create and show all CT's (if verbose) with slider applet
CT_list = create_CT_tables(
df, L_dev, k) # create all combinations of Contingency Tables
if verbose:
print("No of tables = Choosing 2 from " + str(L_dev.shape[1]) +
" LFs = " + str(L_dev.shape[1]) + "C2 = " + str(len(CT_list)))
#print("Note: Showing subtables where CI is not clearly evident")
interact(show_CT,
q=IntSlider(min=1, max=len(CT_list), value=0, step=1),
CT_list=fixed(CT_list))
def get_conditional_deps(CT_list, sig, k, delta=0):
"""peform 3-way table chi-square independence test and obtain test statistic chi_square
(or corresponding p-value) for each CT table where each CT-table is a ({k+1}^{no of other LFs})x(k+1)x(k+1) matrix
https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.chi2_contingency.html"""
CD_edges = []
CD_nodes = []
CD_edges_p_vals = []
p_vals_sum_dict = {}
CT_reduced_list = []
count = 0
#n_bad = 0
for CT in CT_list:
count += 1
Z = int(len(CT.values) /
3) # no of rows/3 (this is new in the fast version)
CT_reshaped = np.reshape(CT.values, (Z, k + 1, k + 1))
if policy == 'old':
p_val_tot, CT_reduced = get_p_total_old_policy(
CT_reshaped, k, Z, delta, sig, count, verbose,
return_more_info
) # Older policy of one round of 0 row/col reduction and then adding delta
else:
p_val_tot, CT_reduced = get_p_total_new_policy(
CT_reshaped, k, Z, sig, count, verbose,
return_more_info) # newer policy of complete reduction
CT_reduced_list.append(CT_reduced)
# checking if total p_value is lesser than chosen sig
if p_val_tot < sig:
digits_LF1 = CT.index.names[1 + no_other_LFs][
3:] # add 1 for the GT column
digits_LF2 = CT.columns.name[
3:] # 3rd index onwards to remove LF_ (3 characters)
CD_edges.append((int(digits_LF1), int(digits_LF2)))
CD_edges_p_vals.append(p_val_tot)
#printing info
edges_info_dict = {}
edges_info_dict['CD_edges'] = CD_edges
edges_info_dict['CD_edges_p_vals'] = CD_edges_p_vals
if verbose:
edges_df = pd.DataFrame(edges_info_dict)
print(edges_df)
return edges_info_dict, CT_reduced_list
# retrieve modified CTs & tuples of LFs that are conditionally independent
edges_info_dict, CT_reduced_list = get_conditional_deps(CT_list,
sig=sig,
k=k,
delta=1)
# display reduced contingency tables' submatrices whose all elements = delta is not true
if verbose:
non_delta_tuple_indices = []
for q in range(len(CT_reduced_list)):
for r in range(len(CT_reduced_list[q])):
delta = 1
if ~(CT_reduced_list[q][r] == delta).all():
non_delta_tuple_indices.append(
((q, r), (q, r))
) # apending a tuple of tuples because first part of outer tuple is key and second is value passed gy slider to fn
def show_CT_sub_matrices(t, CT_reduced_list):
"""function to return qth CT at index q-1 of CT_list"""
return CT_reduced_list[t[0]][t[1]]
print(
"The reduced and modified contingency tables with non-delta values are given below"
)
interact(
show_CT_sub_matrices,
t=Dropdown(
description='index tuples (table number, submatrix number)',
options=non_delta_tuple_indices),
CT_reduced_list=fixed(CT_reduced_list))
if return_more_info:
return edges_info_dict
else:
return edges_info_dict['CD_edges']
def show_masked_images_sagittal(image, mask):
interact(display_images_with_mask_sagittal,
image_z=(0, image.GetSize()[0] - 1),
fixed=fixed(image),
moving=fixed(mask))
def manual_ps(data, notebook=False):
"""
Manual Phase correction using matplotlib
A matplotlib widget is used to manually correct the phase of a Fourier
transformed dataset. If the dataset has more than 1 dimensions, the first
trace will be picked up for phase correction. Clicking the 'Set Phase'
button will print the current linear phase parameters to the console.
A ipywidget is provided for use with Jupyter Notebook to avoid changing
backends. This can be accessed with notebook=True option in this function
.. note:: Needs matplotlib with an interactive backend.
Parameters
----------
data : ndarray
Array of NMR data.
notebook : Bool
True for plotting interactively in Jupyter Notebook
Uses ipywidgets instead of matplotlib widgets
Returns
-------
p0, p1 : float
Linear phase correction parameters. Zero and first order phase
corrections in degrees calculated from pc0, pc1 and pivot displayed
in the interactive window.
Examples
--------
>>> import nmrglue as ng
>>> p0, p1 = ng.process.proc_autophase.manual_ps(data)
>>> # do manual phase correction and close window
>>> phased_data = ng.proc_base.ps(data, p0=p0, p1=p1)
If you are using a Jupyter Notebook::
In [1] ng.process.proc_autophase.manual_ps(data)
Out [1] # do manual phase correction. p0 and p1 values will be updated
# continuously as you do so and are printed below the plot
In [2] phased_data = ng.proc_base.ps(data, p0=p0, p1=p1)
"""
if len(data.shape) == 2:
data = data[0, ...]
elif len(data.shape) == 3:
data = data[0, 0, ...]
elif len(data.shape) == 4:
data = data[0, 0, 0, ...]
if notebook:
from ipywidgets import interact, fixed
import matplotlib.pyplot as plt
def phasecorr(dataset, phcorr0, phcorr1, pivot):
fig, ax = plt.subplots(figsize=(10, 7))
phaseddata = dataset * np.exp(
1j * (phcorr0 + phcorr1 * (
np.arange(-pivot, -pivot+dataset.size)/dataset.size)))
ax.plot(np.real(phaseddata))
ax.set(ylim=(np.min(np.real(data))*2, np.max(np.real(data))*2))
ax.axvline(pivot, color='r', alpha=0.5)
plt.show()
p0 = np.round(
(phcorr0 - phcorr1 * pivot/dataset.size) * 360 / 2 / np.pi, 3)
p1 = np.round(phcorr1*360/2/np.pi, 3)
print('p0 =', p0, 'p1 =', p1)
interact(
phasecorr,
dataset=fixed(data),
phcorr0=(-np.pi, np.pi, 0.01),
phcorr1=(-10*np.pi, 10*np.pi, 0.01),
pivot=(0, data.size, 1))
else:
from matplotlib.widgets import Slider, Button
import matplotlib.pyplot as plt
plt.subplots_adjust(left=0.25, bottom=0.35)
interactive, = plt.plot(data.real, lw=1, color='black')
axcolor = 'white'
axpc0 = plt.axes([0.25, 0.10, 0.65, 0.03], facecolor=axcolor)
axpc1 = plt.axes([0.25, 0.15, 0.65, 0.03], facecolor=axcolor)
axpiv = plt.axes([0.25, 0.20, 0.65, 0.03], facecolor=axcolor)
axpst = plt.axes([0.25, 0.25, 0.15, 0.04], facecolor=axcolor)
spc0 = Slider(axpc0, 'p0', -360, 360, valinit=0)
spc1 = Slider(axpc1, 'p1', -360, 360, valinit=0)
spiv = Slider(axpiv, 'pivot', 0, data.size, valinit=0)
axps = Button(axpst, 'Set Phase', color=axcolor)
def update(val):
pc0 = spc0.val * np.pi / 180
pc1 = spc1.val * np.pi / 180
pivot = spiv.val
interactive.set_ydata((data * np.exp(
1.0j * (pc0 + (pc1 * np.arange(-pivot, -pivot + data.size) /
data.size))).astype(data.dtype)).real)
plt.draw()
def setphase(val):
p0 = spc0.val-spc1.val*spiv.val/data.size
p1 = spc1.val
print(p0, p1)
spc0.on_changed(update)
spc1.on_changed(update)
spiv.on_changed(update)
axps.on_clicked(setphase)
plt.show(block=True)
p0 = spc0.val-spc1.val*spiv.val/data.size
p1 = spc1.val
return p0, p1
#!/usr/bin/env python3
import matplotlib.pyplot as plt
import numpy as np
from ipywidgets import interact, fixed
def optimaler_winkel(v_0, z_0, g=9.81):
return np.rad2deg(np.arcsin(v_0/np.sqrt(2 * (v_0**2 + z_0 * g))))
def bahnkurve(alpha=45, v_0=13.5, z_0=0, g=9.81):
""" alpha - Wurfwinkel [°],
v_0 - Anfangsgeschwindigkeit [m/s],
z_0 - Anfangshöhe [m],
g - Fallbeschleunigung [m/s²]
"""
x = np.arange(0, 25, 0.1) # horizontale Koordinate [m]
print("Optimaler Wurfwinkel: {:4.2f}".format(optimaler_winkel(v_0, z_0, g)))
tan_alpha = np.tan(np.deg2rad(alpha))
z = -g * (1 + tan_alpha**2) / 2 / v_0**2 * x**2 + tan_alpha * x + z_0 # Bahn z(x)
print("Höchster Punkt: z_max = {:4.2f} m bei x = {:4.2f} m".format(z.max(), x[z.argmax()]))
ende = np.where(z < 0)[0][0] # Index der Wurfweite
print("Wurfweite: {:5.2f} m".format(x[ende]))
plt.figure(figsize=[10, 6])
plt.plot(x[:ende], z[:ende])
plt.xlabel("Ort x [m]")
plt.ylabel("Höhe z [m]")
plt.title("Bahnkurve beim Kugelstoßen (EPI/2.3)")
plt.show()
interact(bahnkurve, alpha=(0.0, 90.0 ,1.0), v_0=(1, 15, .5), z_0=(0, 3, .1), g=fixed(9.81));
def interact_drug_data():
from ipywidgets import interact, fixed
# all we need to do is specify here what function we are interacting with
interact(drug_plot, drug=an_epic_party, drug_use=fixed(drug_use))
orientation='horizontal',
style=style,
layout=layout)
D = FloatSlider(min=0,
max=5,
step=0.1,
value=best_D,
description='Disk Prefactor',
readout_format='.2f',
orientation='horizontal',
style=style,
layout=layout)
#rc = FloatSlider(min=0, max=5, step=0.1, value=best_rc, description='Halo Core Radius [kpc]', readout_format='.2f', orientation='horizontal', style=style, layout=layout)
rc = fixed(best_rc)
#rho00 = fixed(best_rho00)
rho00 = FloatSlider(min=0,
max=1e9,
step=1e7,
value=best_rho00,
description=r'Halo Surface Density [$M_{\odot} / pc^3$]',
readout_format='.2e',
orientation='horizontal',
style=style,
layout=layout)
# Interactive widget
def interactive_plot(f):
def plot_design(design):
output_notebook()
sel_mult = SelectMultiple(options = [x for x in design.layers])
interactive_plot = interactive(layer_select, m=sel_mult, design=fixed(design))
display(interactive_plot)
