err = np.zeros([num_trials, 1], dtype='float')
Sn = np.zeros([N, num_dim, num_trials], dtype='float')
#P_new(idx,:), iter(idx,:), err(idx,:)]= calc_3d_tdoa_position(Sn(:,:, idx), Po, v)
for idx in range(0, num_trials):
Sn[:, :, idx] = S + np.random.normal(0, range_err, size=(N, num_dim))
Tn = T + np.random.normal(0, time_err, size=(N))
P_new[idx], iter[idx], err[idx] = calc_tdoa_position(Sn[:, :, idx], Tn, Po, v)
# get the center/means in each direction
cp = np.mean(P_new, axis=0)
r_x, r_y, aou = calc_covariance_matrix(P_new, cp, num_trials)
# results_div.text = tdoa_results + str(aou) + "</font>"
aou_text.value = "{:.6f}".format(aou)
# st_source.data = dict(sx=Sn[:, 0, :].reshape(-1), sy=Sn[:, 1, :].reshape(-1))
etx_source.data = dict(x=P_new[:, 0]*1, y=P_new[:, 1]*1)
ctx_source.data = dict(x=[cp[0]*1], y=[cp[1]*1])
ell_source.data = dict(x=r_x*1, y=r_y*1)
# setup the event callbacks for the plot
st_source.js_on_change('patching', update_plot_callback)
ig_source.js_on_change('patching', update_plot_callback)
tx_source.js_on_change('patching', update_plot_callback)
p_err_spin.js_on_change('value', update_plot_callback)
t_err_spin.js_on_change('value', update_plot_callback)
# p_err_spin.on_change('value', update_plot)
def doc_maker():
'''
make the whole document
'''
global spt_data, custom_path
curdoc().clear() # removes everything in the current document
# dropdown to select a spectrum
select_spectrum = Select(title="Select a spectrum:",
value='',
options=[''] + sorted(os.listdir(spec_path)),
name="select_spectrum",
width=200)
# textinput to give the full path to the location of spectra
path_input = TextInput(title='Spectra folder',
width=200,
name="path_input")
path_input.on_change('value', update_spec_path)
# button to load the spectrum selected in the 'select_spectrum' dropdown
load_button = Button(label='Load spectrum',
width=200,
css_classes=["custom_button"])
load_button.on_click(load_spectrum)
if spt_data == {}:
curdoc().add_root(widgetbox(path_input, select_spectrum, load_button))
if custom_path:
path_input.value = custom_path
else:
path_input.value = spec_path
return
spectrum = spt_data['cur_spec']
header = spt_data[spectrum]['header']
species = np.array([spt_data[spectrum]['columns'][var] for var in header])
SZA = str(spt_data[spectrum]['sza'])
zobs = str(spt_data[spectrum]['zobs'])
freq = species[0] # the frequency list
tm = species[1] # measured transmittance list
tc = species[2] # calculated transmittance list
cont = species[3] # continuum
not_gas = 4 # number of column that are not retrieved species
residuals = spt_data[spectrum]['resid'] # 100*(calculated - measured)
sigma_rms = spt_data[spectrum]['rms_resid'] # sqrt(mean(residuals**2))
# spectrum figure
fig = figure(name="spec_fig",
title=spectrum + '; SZA=' + SZA + '; zobs=' + zobs +
'km; %resid=100*(Measured-Calculated); RMSresid=' +
('%.4f' % sigma_rms) + '%',
plot_width=1000,
plot_height=400,
tools=TOOLS,
y_range=Range1d(-0.04, 1.04),
outline_line_alpha=0,
active_inspect="crosshair",
active_drag="box_zoom")
# residual figure
fig_resid = figure(name="resid_fig",
plot_width=1000,
plot_height=150,
x_range=fig.x_range,
tools=TOOLS,
y_range=Range1d(-3, 3),
outline_line_alpha=0,
active_inspect="crosshair",
active_drag="box_zoom")
# axes labels
fig_resid.xaxis.axis_label = u'Wavenumber (cm\u207B\u00B9)'
fig_resid.yaxis.axis_label = '% Residuals'
fig.yaxis.axis_label = 'Transmittance'
#fig.xaxis.axis_label = u'Wavenumber (cm\u207B\u00B9)'
for elem in [fig, fig_resid]:
elem.yaxis.axis_label_text_font_size = "14pt"
elem.yaxis.major_label_text_font_size = "13pt"
elem.xaxis.axis_label_text_font_size = "14pt"
elem.xaxis.major_label_text_font_size = "13pt"
N_plots = list(
range(len(species) - 2)
) # a range list from 0 to the number of plots, used by the checkbox group
# group of checkboxes that will be used to toggle line and HoverTool visibility
checkbox = CheckboxGroup(labels=header[3:] + ['Measured', 'Calculated'],
active=N_plots,
width=200)
# plotting species lines
plots = []
for j in range(len(species) - not_gas):
try:
plots.append(
fig.line(x=freq,
y=species[j + not_gas],
color=colors[header[j + not_gas]],
line_width=2,
name=header[j + not_gas]))
except KeyError:
print(
'KeyError:', header[j + not_gas],
'is not specified in the "colors" dictionary, you need to add it with an associated color'
)
sys.exit()
# each line has a associated hovertool with a callback that looks at the checkboxes status for the tool visibility.
fig.add_tools(
HoverTool(mode='vline',
line_policy='prev',
renderers=[plots[j]],
names=[header[j + not_gas]],
tooltips=OrderedDict([('name', header[j + not_gas]),
('index', '$index'),
('(x;y)', '(@x{0.00} ; @y{0.000})')
])))
# adding the measured spectrum
plots.append(fig.line(x=freq, y=tm, color='black', line_width=2,
name='Tm'))
fig.add_tools(
HoverTool(mode='vline',
line_policy='prev',
renderers=[plots[j + 1]],
names=['Tm'],
tooltips=OrderedDict([('name', 'Measured'),
('index', '$index'),
('(x;y)', '(@x{0.00} ; @y{0.000})')])))
# adding the calculated spectrum
plots.append(
fig.line(x=freq, y=tc, color='chartreuse', line_width=2, name='Tc'))
#fig.add_tools( HoverTool(mode='vline',line_policy='prev',renderers=[plots[j+2]],names=['Tc'],tooltips=OrderedDict( [('name','Calculated'),('index','$index'),('(x;y)','(@x{0.00} ; @y{0.000})')] )) )
# adding the continuum
#plots.append(fig.line(x=freq,y=cont,color='#FF3399',line_dash='dashed',line_width=2,name='Cont'))
#fig.add_tools( HoverTool(mode='vline',line_policy='prev',renderers=[plots[j+1]],names=['Cont'],tooltips=OrderedDict( [('name','Continuum'),('index','$index'),('(x;y)','(@x{0.00} ; @y{0.000})')] )) )
# legend outside of the figure
fig_legend = Legend(items=[(header[j + not_gas], [plots[j]])
for j in range(len(species) - not_gas)] +
[('Measured', [plots[-2]]),
('Calculated', [plots[-1]])],
location=(0, 0),
border_line_alpha=0)
fig.add_layout(fig_legend, 'right')
fig.legend.click_policy = "hide"
fig.legend.inactive_fill_alpha = 0.6
# now the residual figure
fig_resid.line(x=freq, y=residuals, color='black', name='residuals')
fig_resid.line(x=freq, y=np.zeros(len(freq)), color='red')
fig_resid.add_tools(
HoverTool(mode='vline',
line_policy='prev',
names=['residuals'],
tooltips={
'index': '$index',
'(x;y)': '($x{0.00} ; $y{0.000})'
}))
# set up a dummy legend for the residual figure so that it aligns with the spectrum figure
dummy = fig_resid.line(x=freq,
y=[0 for i in range(len(freq))],
color='white',
visible=False,
alpha=0)
fig_resid_legend = Legend(items=[(' ', [dummy])],
location=(0, 0),
border_line_alpha=0)
fig_resid.add_layout(fig_resid_legend, 'right')
# checkbox group callback
checkbox_iterable = [('p' + str(i), plots[i])
for i in N_plots] + [('checkbox', checkbox)]
checkbox_code = ''.join([
'p' + str(i) + '.visible = checkbox.active.includes(' + str(i) + ');'
for i in N_plots
])
checkbox.callback = CustomJS(
args={key: value
for key, value in checkbox_iterable},
code=checkbox_code)
# button to uncheck all checkboxes
clear_button = Button(label='Hide all lines', width=200)
clear_button_code = """checkbox.active=[];""" + checkbox_code
clear_button.callback = CustomJS(
args={key: value
for key, value in checkbox_iterable},
code=clear_button_code)
# button to check all checkboxes
check_button = Button(label='Show all lines', width=200)
check_button_code = """checkbox.active=""" + str(
N_plots) + """;""" + checkbox_code
check_button.callback = CustomJS(
args={key: value
for key, value in checkbox_iterable},
code=check_button_code)
# extension for the saved file name based on the path to spectra
ext = custom_path.split(os.sep)[-2]
# title div
div = Div(
text='<p align="center"><font size=4><b>{}</b></font></p>'.format(ext),
width=fig.plot_width - 100)
add_vlinked_crosshairs(fig, fig_resid)
sub_grid = gridplot([[fig], [fig_resid], [div]], toolbar_location="left")
# button to activate/deactivate hover tools
hover_button = Button(label='Enable hover tools',
button_type='success',
width=200)
hover_list = [i for i in sub_grid.select({"type": HoverTool})]
# in 'comp' mode, each plot has a different hover tool, I hide them and this button will click them all at once.
hover_button_code = """
if(cb_obj.button_type.includes("success")){
cb_obj.button_type = 'warning';
cb_obj.label = 'Disable hover tools'
} else {
cb_obj.button_type = 'success';
cb_obj.label= 'Enable hover tools';
}
""" + ''.join([
"hover{}.active = !hover{}.active;".format(i, i)
for i in range(len(hover_list))
])
hover_button.callback = CustomJS(
args={'hover{}'.format(i): elem
for i, elem in enumerate(hover_list)},
code=hover_button_code)
# put all the widgets in a box
group = widgetbox(clear_button,
check_button,
hover_button,
width=200,
name="group")
# the final grid for static plots
grid = gridplot([[sub_grid, group]], toolbar_location=None)
# save a standalone html document in spectra_app/save
with open(os.path.join(save_path, '{}_{}.html'.format(spectrum, ext)),
'w') as outfile:
outfile.write(file_html(grid, CDN, spectrum[:12] + spectrum[-3:]))
group = widgetbox(clear_button,
check_button,
hover_button,
path_input,
select_spectrum,
load_button,
width=200)
app_grid = gridplot([[sub_grid, group]], toolbar_location=None)
# add that grid to the document
curdoc().add_root(app_grid)
if custom_path:
path_input.value = custom_path
else:
path_input.value = spec_path
