def _add_multiselect(self):
self.multiselect = MultiSelect(title='States:', value=['01'],
options=self.cases.options)
self.multiselect.max_width = 170
self.multiselect.min_height = 500 - 47
self.multiselect.on_change('value', self._callback_cases)
self.multiselect.on_change('value', self._callback_deaths)
def create_multiselect(cont):
"""
Create a multiselect box for each continent, for the "All" tab.
Args:
cont (str): Continent name.
Returns:
multi_select (:obj:`MultiSelect`): Multiselect box.
"""
multi_select = MultiSelect(title=cont, value=[], options=by_cont[cont])
multi_select.on_change("value", multi_update)
return multi_select
def layout_hook(self):
""" Hook for layout creation. """
options = [(v, v) for v in np.unique(self.context.data[self.coord])]
layout = MultiSelect(title=self.coord,
value=[options[0][0]],
options=options)
layout.size = min(len(options), self.max_elements)
layout.on_change("value", self.on_selected_coord_change)
self.coord_vals = [options[0][0]]
self.context._update_handlers()
return layout
def document(doc):
sliders = []
pre = PreText(text='something')
select = MultiSelect(title='Select the columns',
options=list(df.columns))
button = Button(label='Update')
for col in cols:
MIN, MAX = df[col].min(), df[col].max()
STEP = (MAX - MIN) / 100
slider = RangeSlider(start=MIN,
end=MAX,
step=STEP,
value=(MIN, MAX),
title=col)
sliders.append(slider)
def update():
values = []
txt = '({col} > {low}) & ({col} < {high})'
for slider in sliders:
low, high = slider.value
low, high = float(low), float(high)
formula = txt.format(col=slider.title, low=low, high=high)
values.append(formula)
q = '&'.join(values)
summary_cols = select.value
text = df.query(q)[summary_cols].describe()
pre.text = str(text)
button.on_click(update)
l = layout([column(sliders), [select, button]], pre)
doc.add_root(l)
def update_dataset(attr, old, new):
selected_dataset = dataset.value
rank_slice.end = data_manager.get_size(selected_dataset)
rank_slice.update(value=(1, data_manager.get_size(selected_dataset)))
metadata_type = data_manager.get_metadata_type(selected_dataset)
metadata_domain = data_manager.get_metadata_domain(selected_dataset)
while len(metadata_filters) > 0:
metadata_filters.pop()
for attribute in metadata_type:
m_type = metadata_type[attribute]
m_domain = metadata_domain[attribute]
if m_type == 'boolean':
filter = Select(title=attribute,
value="Any",
options=["Any", "True", "False"])
elif m_type == 'numerical':
filter = RangeSlider(start=m_domain[0],
end=m_domain[1],
value=m_domain,
step=1,
title=attribute)
elif m_type == 'categorical':
categories = sorted(list(metadata_domain[attribute]))
filter = MultiSelect(title=attribute,
value=categories,
options=categories)
elif m_type == 'set':
categories = sorted(list(metadata_domain[attribute]))
filter = MultiSelect(title=attribute,
value=categories,
options=categories)
else:
raise ValueError(
'Unsupported attribute type {} in metadata'.format(m_type))
metadata_filters.append(filter)
for control in metadata_filters:
if hasattr(control, 'value'):
control.on_change('value', update)
if hasattr(control, 'active'):
control.on_change('active', update)
inputs.children = build_controls()
update(attr, old, new)
def __init__(self):
self.rubric = pd.read_excel(
os.path.join(file_path, "data/Rubric.xlsx"), "Rubric v3")
self.cost_model = pd.read_excel(
os.path.join(file_path, "data/Rubric.xlsx"), "Cost_Model")
try:
self.rubric.drop(["Category", "Definition", "Grading Scale"],
inplace=True,
axis=1)
except KeyError:
pass
self.criteria = self.rubric["Criteria"].drop_duplicates().tolist()
self.swing_table = swing_table.create_swing_table()
self.chosen_criteria = []
self.criteria_selection = MultiSelect(title="Choose Criteria:",
size=10)
self.choose_criteria()
self.rubric_values = self.rubric.replace("Excellent", 1.0)
self.rubric_values.replace("Good", 0.5, inplace=True)
self.rubric_values.replace("Poor", 0, inplace=True)
self.rubric_values = self.rubric_values.melt(id_vars=["Criteria"],
var_name=["Tool"],
value_name="Score")
self.weight_sliders = OrderedDict()
self.ranking = OrderedDict()
self.b = Button(label="Update Model", button_type="primary")
self.b.on_click(self.submit_callback)
self.criteria_b = Button(label="Submit Criteria",
button_type="primary")
self.criteria_b.on_click(self.choose_criteria_callback)
self.clear_button = Button(label="Reset", button_type="warning")
self.clear_button.on_click(self.clear_model)
self.rank_submit = Button(label="Calculate Ranks",
button_type="primary")
self.rank_submit.on_click(self.submit_ranks)
self.source = ColumnDataSource()
self.data_table = DataTable
self.app_layout = layout()
def timeplot(data):
#input data is a DataFrame
time = pd.DatetimeIndex(data['ltime'])
#String list to store column names from the third column of the dataframe
columns = []
for x in data.columns[1:]:
columns.append(x)
#change string to float in the data
for x in columns[0:(len(columns) - 2)]:
if (type(data[x][0]) is str):
for i in range(len(data[x])):
data[x][i] = float(data[x][i].replace(',', ''))
output_notebook()
y = data[columns[1]]
x = time
p = Figure(x_axis_type='datetime', title="TimeSeries Plotting")
source = ColumnDataSource(data=dict(x=x, y=y, d=data))
#create a new columndatasoure to pass column name to CustomJS
source2 = ColumnDataSource(data=dict(columns=columns))
p.line('x', 'y', source=source)
p.xaxis.axis_label = "Time"
p.yaxis.axis_label = "Selected Y"
callback = CustomJS(args=dict(source=source, columns=source2),
code="""
var data = source.get('data');
var columns = columns.get('data')['columns'];
var f = cb_obj.get('value');
y = data['y'];
console.log('y');
console.log(y);
var d = data['d'];
//get the index of the chosen column from the widget
for(i = 0; i<columns.length;i++){
if(f[0]==columns[i]){
index = i;
}
}
//make the column transpose since the CustomJS
//takes dataframe as an array of arrays which is
//a row of the DataFrame
for (i = 0; i < d.length; i++) {
y[i] = d[i][index+1];
}
console.log('y');
console.log(y.length);
source.trigger('change');
""")
select = MultiSelect(title="Y_Option:",
value=[columns[0]],
options=columns,
callback=callback)
layout = vform(select, p)
show(layout)
def __init__(
self,
name="Specials",
descr="Choose one",
kind="specials",
css_classes=[],
entries={},
default="",
title=None,
none_allowed=False,
):
self.name = name
self.descr = descr
self.entries = entries
self.kind = kind
self.css_classes = css_classes
options = sorted(entries.keys())
if none_allowed:
options = ["None"] + options
if title is None:
title = "."
css_classes = ["deli-selector", "hide-title"]
else:
css_classes = ["deli-selector"]
self.widget = MultiSelect(
options=options,
value=[default],
# height=150,
size=8,
name="deli-selector",
title=title,
css_classes=css_classes,
)
# HACK: force MultiSelect to only have 1 value selected
def multi_select_hack(attr, old, new):
if len(new) > 1:
self.widget.value = old
self.widget.on_change("value", multi_select_hack)
def modify_doc(self, doc):
self.source = ColumnDataSource(data=dict(top=[], left=[], right=[]))
self.kplot = figure()
self.kplot.y_range.start = 0
self.kplot.quad(top="top",
bottom=0,
left="left",
right="right",
source=self.source)
desc = Div(text=description, sizing_mode="stretch_width")
self.sel_procs = MultiSelect(title="Select MPI.OpenMP",
options=self.procs,
value=["All"])
self.sel_plots = Select(title="Select Plot Type",
options=self.plots,
value=self.plots[0])
self.sel_invokes = MultiSelect(title="Select invokes",
options=self.invokes,
value=["All"])
controls = [self.sel_procs, self.sel_invokes, self.sel_plots]
for control in controls:
control.on_change("value", lambda attr, old, new: self.update())
root = column(
desc,
row(self.sel_procs,
column(self.sel_plots, self.kplot, self.sel_invokes)))
self.update()
doc.add_root(root)
doc.title = "Kernel Timing"
def document(doc):
condition = TextInput(title='Enter your condition')
col = MultiSelect(title='Select the columns', options=list(df.columns))
button = Button(label='Update')
pre = PreText()
pre.text = condition.value
def update():
cond = condition.value
cols = col.value
text = df.query(cond)[cols].describe()
pre.text = str(text)
button.on_click(update)
l = layout([[[condition, button], col], pre])
doc.add_root(l)
def count_clients():
df_clients = pd.read_sql("clients", db.engine)
df_users = pd.read_sql("users", db.engine)
df = pd.merge(
df_clients,
df_users[["id", "username"]],
how="left",
left_on="user_id",
right_on="id",
)
df = df.groupby("username").count()
print(df)
options = list(df.index)
plot = figure(x_range=options)
plot.vbar(x=options, top=df.first_name, width=0.75)
select = MultiSelect(options=options, value=[options[0]])
layout = column(select, plot)
return df, layout
# intialize widgets
save_button = Button(label="Save flagged data", button_type="success")
parameter = Select(title="Parameter",
options=["CTDSAL", "CTDTMP"],
value="CTDSAL")
ref_param = Select(title="Reference", options=["SALNTY"], value="SALNTY")
# ref_param.options = ["foo","bar"] # can dynamically change dropdowns
station = Select(title="Station")
# explanation of flags:
# https://cchdo.github.io/hdo-assets/documentation/manuals/pdf/90_1/chap4.pdf
flag_list = MultiSelect(
title="Plot data flagged as:",
value=["1", "2", "3"],
options=[
("1", "1 [Uncalibrated]"),
("2", "2 [Acceptable]"),
("3", "3 [Questionable]"),
("4", "4 [Bad]"),
],
)
# returns list of select options, e.g., ['2'] or ['1','2']
flag_input = Select(
title="Flag:",
options=[
("1", "1 [Uncalibrated]"),
("2", "2 [Acceptable]"),
("3", "3 [Questionable]"),
("4", "4 [Bad]"),
],
value="3",
)
def plot(tables, output_filename):
'''
This is the plot function that uses Bokeh functions and widgets to make an interactive hexagon plot.
This function recieves:
- tables: dictionary with tables used to create arrays of repeated x, y coordinates (depending on the counts) for the hexagon plot.
- output_filename: filename of .html output in the plots folder
The coordinate arrays are used to create a pandas dataframe with Bokeh functions. This dataframe contains the q, r coordinates and counts used to plot the
hexagons. To this dataframe, extra information is added (e.g. most common chemicals), which is displayed in the hover tooltip.
Gaussian blur is added to copies of this dataframe and given as input to the Bokeh slider widget.
Other widgets are added as well, for saturation, normalisation etc. Bokeh allows to customize these widges with javascript code.
The hexagon plot is saved as a .html file and also shown in the browser.
'''
file_name = 'plots/' + str(output_filename) + '.html'
output_file(file_name)
# Blur and saturation values
BLUR_MAX = 3
BLUR_STEP_SIZE = 1
SATURATION_MAX = 5
SATURATION_STEP_SIZE = 0.25
# First, create array for plot properties ( ratio, size of hexagons etc.)
default_term = list(tables.keys())[0]
x, y, ids = create_array(tables[default_term]['table'],
normalisation=False)
# Hexagon plot properties
length = len(x)
orientation = 'flattop'
ratio = ((max(y) - min(y)) / (max(x) - min(x)))
size = 10 / ratio
h = sqrt(3) * size
h = h * ratio
title = 'Hexbin plot for ' + str(
length) + ' annotated chemicals with query ' + str(default_term)
# make figure
p = figure(title=title,
x_range=[min(x) - 0.5, max(x) + 0.5],
y_range=[0 - (h / 2), max(y) + 100],
tools="wheel_zoom,reset,save",
background_fill_color='#440154')
p.grid.visible = False
p.xaxis.axis_label = "log(P)"
p.yaxis.axis_label = "mass in Da"
p.xaxis.axis_label_text_font_style = 'normal'
p.yaxis.axis_label_text_font_style = 'normal'
# source for plot
term_to_source, term_to_metadata, options = make_plot_sources(
tables, size, ratio, orientation, BLUR_MAX, BLUR_STEP_SIZE)
# start source for plot, this is the source that is first displayed in the hexagon figure
x, y, ids = create_array(tables[default_term]['table'],
normalisation=False)
df = hexbin(x, y, ids, size, aspect_scale=ratio, orientation=orientation)
df = add_counts(df, tables[default_term]['table'])
source = ColumnDataSource(df)
metadata = term_to_metadata[default_term]
metadata = return_html(metadata)
# color mapper
mapper = linear_cmap('scaling', 'Viridis256', 0,
max(source.data['scaling']))
# plot
hex = p.hex_tile(q="q",
r="r",
size=size,
line_color=None,
source=source,
aspect_scale=ratio,
orientation=orientation,
fill_color=mapper)
# HOVER
TOOLTIPS = return_tooltip()
code_callback_hover = return_code('hover')
callback_hover = CustomJS(code=code_callback_hover)
hover = HoverTool(tooltips=TOOLTIPS,
callback=callback_hover,
show_arrow=False)
p.add_tools(hover)
# WIDGETS
slider1 = Slider(start=1,
end=SATURATION_MAX,
value=1,
step=SATURATION_STEP_SIZE,
title="Saturation",
width=100)
slider2 = Slider(start=0,
end=BLUR_MAX,
value=0,
step=BLUR_STEP_SIZE,
title="Blur",
width=100)
checkbox = CheckboxGroup(labels=["TFIDF"], active=[])
radio_button_group = RadioGroup(labels=["Viridis256", "Greys256"],
active=0)
button = Button(label="Metadata", button_type="default", width=100)
multi_select = MultiSelect(title=output_filename,
value=[default_term],
options=options,
width=100,
height=300)
# WIDGETS CODE FOR CALLBACK
code_callback_slider1 = return_code('slider1')
code_callback_slider2 = return_code('slider2')
code_callback_checkbox = return_code('checkbox')
code_callback_rbg = return_code('rbg')
code_callback_button = return_code('button')
code_callback_ms = return_code('multi_select')
# WIDGETS CALLBACK
callback_slider1 = CustomJS(args={
'source': source,
'mapper': mapper
},
code=code_callback_slider1)
callback_slider2 = CustomJS(args={
'source': source,
'mapper': mapper,
'slider1': slider1,
'multi_select': multi_select,
'checkbox': checkbox,
'term_to_source': term_to_source,
'step_size': BLUR_STEP_SIZE
},
code=code_callback_slider2)
callback_checkbox = CustomJS(args={
'source': source,
'term_to_source': term_to_source,
'multi_select': multi_select,
'step_size': BLUR_STEP_SIZE,
'slider1': slider1,
'slider2': slider2,
'mapper': mapper
},
code=code_callback_checkbox)
callback_radio_button_group = CustomJS(args={
'p': p,
'mapper': mapper,
'Viridis256': Viridis256,
'Greys256': Greys256
},
code=code_callback_rbg)
callback_button = CustomJS(args={
'term_to_metadata': term_to_metadata,
'multi_select': multi_select
},
code=code_callback_button)
callback_ms = CustomJS(args={
'source': source,
'term_to_source': term_to_source,
'checkbox': checkbox,
'metadata': metadata,
'step_size': BLUR_STEP_SIZE,
'slider2': slider2,
'slider1': slider1,
'p': p,
'mapper': mapper
},
code=code_callback_ms)
# # WIDGETS INTERACTION
slider1.js_on_change('value', callback_slider1)
slider2.js_on_change('value', callback_slider2)
checkbox.js_on_change('active', callback_checkbox)
radio_button_group.js_on_change('active', callback_radio_button_group)
button.js_on_event(events.ButtonClick, callback_button)
multi_select.js_on_change("value", callback_ms)
# LAYOUT
layout = row(
multi_select, p,
column(slider1, slider2, checkbox, radio_button_group, button))
show(layout)
button_type="primary")
shuffle_button = Button(label="Shake!", button_type="primary")
start_button = Button(label="Start timer", button_type="success")
stop_button = Button(label="Stop timer", button_type="danger")
timer = Div(
text=f"""Timer: <br> 0:00""",
style={
"font-size": "400%",
"color": "black",
"text-align": "center"
},
)
show_words_button = Button(label="Show all words?", button_type="danger")
show_words_options = RadioButtonGroup(labels=["Alphabetical", "By Length"],
active=0)
word_select = MultiSelect(value=[], options=[], height=500, size=10)
word_count = Div(text="")
# make word histogram
hist_src = ColumnDataSource(dict(hist=[], left=[], right=[]))
# pdf_src = ColumnDataSource(dict(x=[], pdf=[]))
p = figure(
plot_height=200,
plot_width=350,
title="Word Count",
tools="",
background_fill_color="#fafafa",
x_axis_label="# of Words",
y_axis_label="# of Games",
)
p.quad(
columns={"New Flag_x": "New Flag", "Comments_x": "Comments"}
).drop(columns=["New Flag_y", "Comments_y"])
# intialize widgets
save_button = Button(label="Save flagged data", button_type="success")
parameter = Select(title="Parameter", options=["CTDSAL", "CTDTMP"], value="CTDSAL")
ref_param = Select(title="Reference", options=["SALNTY"], value="SALNTY")
# ref_param.options = ["foo","bar"] # can dynamically change dropdowns
station = Select(title="Station", options=ssscc_list, value=ssscc_list[0])
# explanation of flags:
# https://cchdo.github.io/hdo-assets/documentation/manuals/pdf/90_1/chap4.pdf
flag_list = MultiSelect(
title="Plot data flagged as:",
value=["1", "2", "3"],
options=[
("1", "1 [Uncalibrated]"),
("2", "2 [Acceptable]"),
("3", "3 [Questionable]"),
("4", "4 [Bad]"),
],
)
# returns list of select options, e.g., ['2'] or ['1','2']
flag_input = Select(
title="Flag:",
options=[
("1", "1 [Uncalibrated]"),
("2", "2 [Acceptable]"),
("3", "3 [Questionable]"),
("4", "4 [Bad]"),
],
value="3",
)
def create():
doc = curdoc()
det_data = {}
cami_meta = {}
def proposal_textinput_callback(_attr, _old, new):
nonlocal cami_meta
proposal = new.strip()
for zebra_proposals_path in pyzebra.ZEBRA_PROPOSALS_PATHS:
proposal_path = os.path.join(zebra_proposals_path, proposal)
if os.path.isdir(proposal_path):
# found it
break
else:
raise ValueError(f"Can not find data for proposal '{proposal}'.")
file_list = []
for file in os.listdir(proposal_path):
if file.endswith(".hdf"):
file_list.append((os.path.join(proposal_path, file), file))
file_select.options = file_list
cami_meta = {}
proposal_textinput = TextInput(title="Proposal number:", width=210)
proposal_textinput.on_change("value", proposal_textinput_callback)
def upload_button_callback(_attr, _old, new):
nonlocal cami_meta
with io.StringIO(base64.b64decode(new).decode()) as file:
cami_meta = pyzebra.parse_h5meta(file)
file_list = cami_meta["filelist"]
file_select.options = [(entry, os.path.basename(entry))
for entry in file_list]
upload_div = Div(text="or upload .cami file:", margin=(5, 5, 0, 5))
upload_button = FileInput(accept=".cami", width=200)
upload_button.on_change("value", upload_button_callback)
def update_image(index=None):
if index is None:
index = index_spinner.value
current_image = det_data["data"][index]
proj_v_line_source.data.update(x=np.arange(0, IMAGE_W) + 0.5,
y=np.mean(current_image, axis=0))
proj_h_line_source.data.update(x=np.mean(current_image, axis=1),
y=np.arange(0, IMAGE_H) + 0.5)
image_source.data.update(
h=[np.zeros((1, 1))],
k=[np.zeros((1, 1))],
l=[np.zeros((1, 1))],
)
image_source.data.update(image=[current_image])
if main_auto_checkbox.active:
im_min = np.min(current_image)
im_max = np.max(current_image)
display_min_spinner.value = im_min
display_max_spinner.value = im_max
image_glyph.color_mapper.low = im_min
image_glyph.color_mapper.high = im_max
if "mf" in det_data:
metadata_table_source.data.update(mf=[det_data["mf"][index]])
else:
metadata_table_source.data.update(mf=[None])
if "temp" in det_data:
metadata_table_source.data.update(temp=[det_data["temp"][index]])
else:
metadata_table_source.data.update(temp=[None])
gamma, nu = calculate_pol(det_data, index)
omega = np.ones((IMAGE_H, IMAGE_W)) * det_data["omega"][index]
image_source.data.update(gamma=[gamma], nu=[nu], omega=[omega])
def update_overview_plot():
h5_data = det_data["data"]
n_im, n_y, n_x = h5_data.shape
overview_x = np.mean(h5_data, axis=1)
overview_y = np.mean(h5_data, axis=2)
overview_plot_x_image_source.data.update(image=[overview_x],
dw=[n_x],
dh=[n_im])
overview_plot_y_image_source.data.update(image=[overview_y],
dw=[n_y],
dh=[n_im])
if proj_auto_checkbox.active:
im_min = min(np.min(overview_x), np.min(overview_y))
im_max = max(np.max(overview_x), np.max(overview_y))
proj_display_min_spinner.value = im_min
proj_display_max_spinner.value = im_max
overview_plot_x_image_glyph.color_mapper.low = im_min
overview_plot_y_image_glyph.color_mapper.low = im_min
overview_plot_x_image_glyph.color_mapper.high = im_max
overview_plot_y_image_glyph.color_mapper.high = im_max
frame_range.start = 0
frame_range.end = n_im
frame_range.reset_start = 0
frame_range.reset_end = n_im
frame_range.bounds = (0, n_im)
scan_motor = det_data["scan_motor"]
overview_plot_y.axis[1].axis_label = f"Scanning motor, {scan_motor}"
var = det_data[scan_motor]
var_start = var[0]
var_end = var[-1] + (var[-1] - var[0]) / (n_im - 1)
scanning_motor_range.start = var_start
scanning_motor_range.end = var_end
scanning_motor_range.reset_start = var_start
scanning_motor_range.reset_end = var_end
# handle both, ascending and descending sequences
scanning_motor_range.bounds = (min(var_start,
var_end), max(var_start, var_end))
def file_select_callback(_attr, old, new):
nonlocal det_data
if not new:
# skip empty selections
return
# Avoid selection of multiple indicies (via Shift+Click or Ctrl+Click)
if len(new) > 1:
# drop selection to the previous one
file_select.value = old
return
if len(old) > 1:
# skip unnecessary update caused by selection drop
return
det_data = pyzebra.read_detector_data(new[0])
if cami_meta and "crystal" in cami_meta:
det_data["ub"] = cami_meta["crystal"]["UB"]
index_spinner.value = 0
index_spinner.high = det_data["data"].shape[0] - 1
index_slider.end = det_data["data"].shape[0] - 1
zebra_mode = det_data["zebra_mode"]
if zebra_mode == "nb":
metadata_table_source.data.update(geom=["normal beam"])
else: # zebra_mode == "bi"
metadata_table_source.data.update(geom=["bisecting"])
update_image(0)
update_overview_plot()
file_select = MultiSelect(title="Available .hdf files:",
width=210,
height=250)
file_select.on_change("value", file_select_callback)
def index_callback(_attr, _old, new):
update_image(new)
index_slider = Slider(value=0, start=0, end=1, show_value=False, width=400)
index_spinner = Spinner(title="Image index:", value=0, low=0, width=100)
index_spinner.on_change("value", index_callback)
index_slider.js_link("value_throttled", index_spinner, "value")
index_spinner.js_link("value", index_slider, "value")
plot = Plot(
x_range=Range1d(0, IMAGE_W, bounds=(0, IMAGE_W)),
y_range=Range1d(0, IMAGE_H, bounds=(0, IMAGE_H)),
plot_height=IMAGE_PLOT_H,
plot_width=IMAGE_PLOT_W,
toolbar_location="left",
)
# ---- tools
plot.toolbar.logo = None
# ---- axes
plot.add_layout(LinearAxis(), place="above")
plot.add_layout(LinearAxis(major_label_orientation="vertical"),
place="right")
# ---- grid lines
plot.add_layout(Grid(dimension=0, ticker=BasicTicker()))
plot.add_layout(Grid(dimension=1, ticker=BasicTicker()))
# ---- rgba image glyph
image_source = ColumnDataSource(
dict(
image=[np.zeros((IMAGE_H, IMAGE_W), dtype="float32")],
h=[np.zeros((1, 1))],
k=[np.zeros((1, 1))],
l=[np.zeros((1, 1))],
gamma=[np.zeros((1, 1))],
nu=[np.zeros((1, 1))],
omega=[np.zeros((1, 1))],
x=[0],
y=[0],
dw=[IMAGE_W],
dh=[IMAGE_H],
))
h_glyph = Image(image="h", x="x", y="y", dw="dw", dh="dh", global_alpha=0)
k_glyph = Image(image="k", x="x", y="y", dw="dw", dh="dh", global_alpha=0)
l_glyph = Image(image="l", x="x", y="y", dw="dw", dh="dh", global_alpha=0)
gamma_glyph = Image(image="gamma",
x="x",
y="y",
dw="dw",
dh="dh",
global_alpha=0)
nu_glyph = Image(image="nu",
x="x",
y="y",
dw="dw",
dh="dh",
global_alpha=0)
omega_glyph = Image(image="omega",
x="x",
y="y",
dw="dw",
dh="dh",
global_alpha=0)
plot.add_glyph(image_source, h_glyph)
plot.add_glyph(image_source, k_glyph)
plot.add_glyph(image_source, l_glyph)
plot.add_glyph(image_source, gamma_glyph)
plot.add_glyph(image_source, nu_glyph)
plot.add_glyph(image_source, omega_glyph)
image_glyph = Image(image="image", x="x", y="y", dw="dw", dh="dh")
plot.add_glyph(image_source, image_glyph, name="image_glyph")
# ---- projections
proj_v = Plot(
x_range=plot.x_range,
y_range=DataRange1d(),
plot_height=150,
plot_width=IMAGE_PLOT_W,
toolbar_location=None,
)
proj_v.add_layout(LinearAxis(major_label_orientation="vertical"),
place="right")
proj_v.add_layout(LinearAxis(major_label_text_font_size="0pt"),
place="below")
proj_v.add_layout(Grid(dimension=0, ticker=BasicTicker()))
proj_v.add_layout(Grid(dimension=1, ticker=BasicTicker()))
proj_v_line_source = ColumnDataSource(dict(x=[], y=[]))
proj_v.add_glyph(proj_v_line_source,
Line(x="x", y="y", line_color="steelblue"))
proj_h = Plot(
x_range=DataRange1d(),
y_range=plot.y_range,
plot_height=IMAGE_PLOT_H,
plot_width=150,
toolbar_location=None,
)
proj_h.add_layout(LinearAxis(), place="above")
proj_h.add_layout(LinearAxis(major_label_text_font_size="0pt"),
place="left")
proj_h.add_layout(Grid(dimension=0, ticker=BasicTicker()))
proj_h.add_layout(Grid(dimension=1, ticker=BasicTicker()))
proj_h_line_source = ColumnDataSource(dict(x=[], y=[]))
proj_h.add_glyph(proj_h_line_source,
Line(x="x", y="y", line_color="steelblue"))
# add tools
hovertool = HoverTool(tooltips=[
("intensity", "@image"),
("gamma", "@gamma"),
("nu", "@nu"),
("omega", "@omega"),
("h", "@h"),
("k", "@k"),
("l", "@l"),
])
box_edit_source = ColumnDataSource(dict(x=[], y=[], width=[], height=[]))
box_edit_glyph = Rect(x="x",
y="y",
width="width",
height="height",
fill_alpha=0,
line_color="red")
box_edit_renderer = plot.add_glyph(box_edit_source, box_edit_glyph)
boxedittool = BoxEditTool(renderers=[box_edit_renderer], num_objects=1)
def box_edit_callback(_attr, _old, new):
if new["x"]:
h5_data = det_data["data"]
x_val = np.arange(h5_data.shape[0])
left = int(np.floor(new["x"][0]))
right = int(np.ceil(new["x"][0] + new["width"][0]))
bottom = int(np.floor(new["y"][0]))
top = int(np.ceil(new["y"][0] + new["height"][0]))
y_val = np.sum(h5_data[:, bottom:top, left:right], axis=(1, 2))
else:
x_val = []
y_val = []
roi_avg_plot_line_source.data.update(x=x_val, y=y_val)
box_edit_source.on_change("data", box_edit_callback)
wheelzoomtool = WheelZoomTool(maintain_focus=False)
plot.add_tools(
PanTool(),
BoxZoomTool(),
wheelzoomtool,
ResetTool(),
hovertool,
boxedittool,
)
plot.toolbar.active_scroll = wheelzoomtool
# shared frame ranges
frame_range = Range1d(0, 1, bounds=(0, 1))
scanning_motor_range = Range1d(0, 1, bounds=(0, 1))
det_x_range = Range1d(0, IMAGE_W, bounds=(0, IMAGE_W))
overview_plot_x = Plot(
title=Title(text="Projections on X-axis"),
x_range=det_x_range,
y_range=frame_range,
extra_y_ranges={"scanning_motor": scanning_motor_range},
plot_height=400,
plot_width=IMAGE_PLOT_W - 3,
)
# ---- tools
wheelzoomtool = WheelZoomTool(maintain_focus=False)
overview_plot_x.toolbar.logo = None
overview_plot_x.add_tools(
PanTool(),
BoxZoomTool(),
wheelzoomtool,
ResetTool(),
)
overview_plot_x.toolbar.active_scroll = wheelzoomtool
# ---- axes
overview_plot_x.add_layout(LinearAxis(axis_label="Coordinate X, pix"),
place="below")
overview_plot_x.add_layout(LinearAxis(axis_label="Frame",
major_label_orientation="vertical"),
place="left")
# ---- grid lines
overview_plot_x.add_layout(Grid(dimension=0, ticker=BasicTicker()))
overview_plot_x.add_layout(Grid(dimension=1, ticker=BasicTicker()))
# ---- rgba image glyph
overview_plot_x_image_source = ColumnDataSource(
dict(image=[np.zeros((1, 1), dtype="float32")],
x=[0],
y=[0],
dw=[IMAGE_W],
dh=[1]))
overview_plot_x_image_glyph = Image(image="image",
x="x",
y="y",
dw="dw",
dh="dh")
overview_plot_x.add_glyph(overview_plot_x_image_source,
overview_plot_x_image_glyph,
name="image_glyph")
det_y_range = Range1d(0, IMAGE_H, bounds=(0, IMAGE_H))
overview_plot_y = Plot(
title=Title(text="Projections on Y-axis"),
x_range=det_y_range,
y_range=frame_range,
extra_y_ranges={"scanning_motor": scanning_motor_range},
plot_height=400,
plot_width=IMAGE_PLOT_H + 22,
)
# ---- tools
wheelzoomtool = WheelZoomTool(maintain_focus=False)
overview_plot_y.toolbar.logo = None
overview_plot_y.add_tools(
PanTool(),
BoxZoomTool(),
wheelzoomtool,
ResetTool(),
)
overview_plot_y.toolbar.active_scroll = wheelzoomtool
# ---- axes
overview_plot_y.add_layout(LinearAxis(axis_label="Coordinate Y, pix"),
place="below")
overview_plot_y.add_layout(
LinearAxis(
y_range_name="scanning_motor",
axis_label="Scanning motor",
major_label_orientation="vertical",
),
place="right",
)
# ---- grid lines
overview_plot_y.add_layout(Grid(dimension=0, ticker=BasicTicker()))
overview_plot_y.add_layout(Grid(dimension=1, ticker=BasicTicker()))
# ---- rgba image glyph
overview_plot_y_image_source = ColumnDataSource(
dict(image=[np.zeros((1, 1), dtype="float32")],
x=[0],
y=[0],
dw=[IMAGE_H],
dh=[1]))
overview_plot_y_image_glyph = Image(image="image",
x="x",
y="y",
dw="dw",
dh="dh")
overview_plot_y.add_glyph(overview_plot_y_image_source,
overview_plot_y_image_glyph,
name="image_glyph")
roi_avg_plot = Plot(
x_range=DataRange1d(),
y_range=DataRange1d(),
plot_height=150,
plot_width=IMAGE_PLOT_W,
toolbar_location="left",
)
# ---- tools
roi_avg_plot.toolbar.logo = None
# ---- axes
roi_avg_plot.add_layout(LinearAxis(), place="below")
roi_avg_plot.add_layout(LinearAxis(major_label_orientation="vertical"),
place="left")
# ---- grid lines
roi_avg_plot.add_layout(Grid(dimension=0, ticker=BasicTicker()))
roi_avg_plot.add_layout(Grid(dimension=1, ticker=BasicTicker()))
roi_avg_plot_line_source = ColumnDataSource(dict(x=[], y=[]))
roi_avg_plot.add_glyph(roi_avg_plot_line_source,
Line(x="x", y="y", line_color="steelblue"))
cmap_dict = {
"gray": Greys256,
"gray_reversed": Greys256[::-1],
"plasma": Plasma256,
"cividis": Cividis256,
}
def colormap_callback(_attr, _old, new):
image_glyph.color_mapper = LinearColorMapper(palette=cmap_dict[new])
overview_plot_x_image_glyph.color_mapper = LinearColorMapper(
palette=cmap_dict[new])
overview_plot_y_image_glyph.color_mapper = LinearColorMapper(
palette=cmap_dict[new])
colormap = Select(title="Colormap:",
options=list(cmap_dict.keys()),
width=210)
colormap.on_change("value", colormap_callback)
colormap.value = "plasma"
STEP = 1
def main_auto_checkbox_callback(state):
if state:
display_min_spinner.disabled = True
display_max_spinner.disabled = True
else:
display_min_spinner.disabled = False
display_max_spinner.disabled = False
update_image()
main_auto_checkbox = CheckboxGroup(labels=["Main Auto Range"],
active=[0],
width=145,
margin=[10, 5, 0, 5])
main_auto_checkbox.on_click(main_auto_checkbox_callback)
def display_max_spinner_callback(_attr, _old_value, new_value):
display_min_spinner.high = new_value - STEP
image_glyph.color_mapper.high = new_value
display_max_spinner = Spinner(
low=0 + STEP,
value=1,
step=STEP,
disabled=bool(main_auto_checkbox.active),
width=100,
height=31,
)
display_max_spinner.on_change("value", display_max_spinner_callback)
def display_min_spinner_callback(_attr, _old_value, new_value):
display_max_spinner.low = new_value + STEP
image_glyph.color_mapper.low = new_value
display_min_spinner = Spinner(
low=0,
high=1 - STEP,
value=0,
step=STEP,
disabled=bool(main_auto_checkbox.active),
width=100,
height=31,
)
display_min_spinner.on_change("value", display_min_spinner_callback)
PROJ_STEP = 0.1
def proj_auto_checkbox_callback(state):
if state:
proj_display_min_spinner.disabled = True
proj_display_max_spinner.disabled = True
else:
proj_display_min_spinner.disabled = False
proj_display_max_spinner.disabled = False
update_overview_plot()
proj_auto_checkbox = CheckboxGroup(labels=["Projections Auto Range"],
active=[0],
width=145,
margin=[10, 5, 0, 5])
proj_auto_checkbox.on_click(proj_auto_checkbox_callback)
def proj_display_max_spinner_callback(_attr, _old_value, new_value):
proj_display_min_spinner.high = new_value - PROJ_STEP
overview_plot_x_image_glyph.color_mapper.high = new_value
overview_plot_y_image_glyph.color_mapper.high = new_value
proj_display_max_spinner = Spinner(
low=0 + PROJ_STEP,
value=1,
step=PROJ_STEP,
disabled=bool(proj_auto_checkbox.active),
width=100,
height=31,
)
proj_display_max_spinner.on_change("value",
proj_display_max_spinner_callback)
def proj_display_min_spinner_callback(_attr, _old_value, new_value):
proj_display_max_spinner.low = new_value + PROJ_STEP
overview_plot_x_image_glyph.color_mapper.low = new_value
overview_plot_y_image_glyph.color_mapper.low = new_value
proj_display_min_spinner = Spinner(
low=0,
high=1 - PROJ_STEP,
value=0,
step=PROJ_STEP,
disabled=bool(proj_auto_checkbox.active),
width=100,
height=31,
)
proj_display_min_spinner.on_change("value",
proj_display_min_spinner_callback)
def hkl_button_callback():
index = index_spinner.value
h, k, l = calculate_hkl(det_data, index)
image_source.data.update(h=[h], k=[k], l=[l])
hkl_button = Button(label="Calculate hkl (slow)", width=210)
hkl_button.on_click(hkl_button_callback)
def events_list_callback(_attr, _old, new):
doc.events_list_spind.value = new
events_list = TextAreaInput(rows=7, width=830)
events_list.on_change("value", events_list_callback)
doc.events_list_hdf_viewer = events_list
def add_event_button_callback():
diff_vec = []
p0 = [1.0, 0.0, 1.0]
maxfev = 100000
wave = det_data["wave"]
ddist = det_data["ddist"]
gamma = det_data["gamma"][0]
omega = det_data["omega"][0]
nu = det_data["nu"][0]
chi = det_data["chi"][0]
phi = det_data["phi"][0]
scan_motor = det_data["scan_motor"]
var_angle = det_data[scan_motor]
x0 = int(np.floor(det_x_range.start))
xN = int(np.ceil(det_x_range.end))
y0 = int(np.floor(det_y_range.start))
yN = int(np.ceil(det_y_range.end))
fr0 = int(np.floor(frame_range.start))
frN = int(np.ceil(frame_range.end))
data_roi = det_data["data"][fr0:frN, y0:yN, x0:xN]
cnts = np.sum(data_roi, axis=(1, 2))
coeff, _ = curve_fit(gauss,
range(len(cnts)),
cnts,
p0=p0,
maxfev=maxfev)
m = cnts.mean()
sd = cnts.std()
snr_cnts = np.where(sd == 0, 0, m / sd)
frC = fr0 + coeff[1]
var_F = var_angle[math.floor(frC)]
var_C = var_angle[math.ceil(frC)]
frStep = frC - math.floor(frC)
var_step = var_C - var_F
var_p = var_F + var_step * frStep
if scan_motor == "gamma":
gamma = var_p
elif scan_motor == "omega":
omega = var_p
elif scan_motor == "nu":
nu = var_p
elif scan_motor == "chi":
chi = var_p
elif scan_motor == "phi":
phi = var_p
intensity = coeff[1] * abs(
coeff[2] * var_step) * math.sqrt(2) * math.sqrt(np.pi)
projX = np.sum(data_roi, axis=(0, 1))
coeff, _ = curve_fit(gauss,
range(len(projX)),
projX,
p0=p0,
maxfev=maxfev)
x_pos = x0 + coeff[1]
projY = np.sum(data_roi, axis=(0, 2))
coeff, _ = curve_fit(gauss,
range(len(projY)),
projY,
p0=p0,
maxfev=maxfev)
y_pos = y0 + coeff[1]
ga, nu = pyzebra.det2pol(ddist, gamma, nu, x_pos, y_pos)
diff_vector = pyzebra.z1frmd(wave, ga, omega, chi, phi, nu)
d_spacing = float(pyzebra.dandth(wave, diff_vector)[0])
diff_vector = diff_vector.flatten() * 1e10
dv1, dv2, dv3 = diff_vector
diff_vec.append(diff_vector)
if events_list.value and not events_list.value.endswith("\n"):
events_list.value = events_list.value + "\n"
events_list.value = (
events_list.value +
f"{x_pos} {y_pos} {intensity} {snr_cnts} {dv1} {dv2} {dv3} {d_spacing}"
)
add_event_button = Button(label="Add spind event")
add_event_button.on_click(add_event_button_callback)
metadata_table_source = ColumnDataSource(
dict(geom=[""], temp=[None], mf=[None]))
num_formatter = NumberFormatter(format="0.00", nan_format="")
metadata_table = DataTable(
source=metadata_table_source,
columns=[
TableColumn(field="geom", title="Geometry", width=100),
TableColumn(field="temp",
title="Temperature",
formatter=num_formatter,
width=100),
TableColumn(field="mf",
title="Magnetic Field",
formatter=num_formatter,
width=100),
],
width=300,
height=50,
autosize_mode="none",
index_position=None,
)
# Final layout
import_layout = column(proposal_textinput, upload_div, upload_button,
file_select)
layout_image = column(
gridplot([[proj_v, None], [plot, proj_h]], merge_tools=False))
colormap_layout = column(
colormap,
main_auto_checkbox,
row(display_min_spinner, display_max_spinner),
proj_auto_checkbox,
row(proj_display_min_spinner, proj_display_max_spinner),
)
layout_controls = column(
row(metadata_table, index_spinner,
column(Spacer(height=25), index_slider)),
row(add_event_button, hkl_button),
row(events_list),
)
layout_overview = column(
gridplot(
[[overview_plot_x, overview_plot_y]],
toolbar_options=dict(logo=None),
merge_tools=True,
toolbar_location="left",
), )
tab_layout = row(
column(import_layout, colormap_layout),
column(layout_overview, layout_controls),
column(roi_avg_plot, layout_image),
)
return Panel(child=tab_layout, title="hdf viewer")
u'primary visual cortex (striate cortex, area V1/17)',
u'striatum',
u'primary motor cortex (area M1, area 4)',
u'posteroventral (inferior) parietal cortex',
u'primary somatosensory cortex (area S1, areas 3,1,2)',
u'cerebellum',
u'cerebellar cortex',
u'mediodorsal nucleus of thalamus']
df1, df2 = get_dataframes(gene, structures)
source1 = ColumnDataSource(df1)
source2 = ColumnDataSource(df2)
source3, source4 = get_boxplot_data(df1)
age_plot = expression_by_age(source1, source2)
structure_plot = structure_boxplot(source3, source4)
source5 = get_heatmap_data(genes1)
heatmap = plot_heatmap(source5, genes1)
plot = vplot(structure_plot,age_plot, heatmap)
multi_select = MultiSelect(title="Brain Regions:", value=structures,
options=structures)
multi_select.on_change('value', update_plot)
select = Select(title="Gene:", value=genes[2], options=genes)
select.on_change('value', update_plot)
curdoc().add_root(vplot(select, plot, multi_select))
def bkapp(doc):
### Functions ###
# functions for user dialogs
def open_file(ftype):
root = Tk()
root.withdraw()
file = askopenfilename(filetypes=ftype,
title='Open File',
initialdir=os.getcwd())
root.destroy()
return file
def choose_directory():
root = Tk()
root.withdraw()
out_dir = askdirectory()
root.destroy()
return out_dir
def write_output_directory(output_dir):
root = Tk()
root.withdraw()
makeDir = askquestion('Make Directory','Output directory not set. Make directory: '
+output_dir+'? If not, you\'ll be prompted to change directories.',icon = 'warning')
root.destroy()
return makeDir
def overwrite_file():
root = Tk()
root.withdraw()
overwrite = askquestion('Overwrite File','File already exits. Do you want to overwrite?',icon = 'warning')
root.destroy()
return overwrite
def update_filename():
filetype = [("Video files", "*.mp4")]
fname = open_file(filetype)
if fname:
#print('Successfully loaded file: '+fname)
load_data(filename=fname)
def change_directory():
out_dir = choose_directory()
if out_dir:
source.data["output_dir"] = [out_dir]
outDir.text = out_dir
return out_dir
# load data from file
def load_data(filename):
vidcap = cv2.VideoCapture(filename)
success,frame = vidcap.read()
img_tmp,_,__ = cv2.split(frame)
h,w = np.shape(img_tmp)
img = np.flipud(img_tmp)
radio_button_gp.active = 0
fname = os.path.split(filename)[1]
input_dir = os.path.split(filename)[0]
if source.data['output_dir'][0]=='':
output_dir = os.path.join(input_dir,fname.split('.')[0])
else:
output_dir = source.data['output_dir'][0]
if not os.path.isdir(output_dir):
makeDir = write_output_directory(output_dir)
if makeDir=='yes':
os.mkdir(output_dir)
else:
output_dir = change_directory()
if output_dir:
source.data = dict(image_orig=[img], image=[img], bin_img=[0],
x=[0], y=[0], dw=[w], dh=[h], num_contours=[0], roi_coords=[0],
img_name=[fname], input_dir=[input_dir], output_dir=[output_dir])
curr_img = p.select_one({'name':'image'})
if curr_img:
p.renderers.remove(curr_img)
p.image(source=source, image='image', x='x', y='y', dw='dw', dh='dh', color_mapper=cmap,level='image',name='image')
p.plot_height=int(h/2)
p.plot_width=int(w/2)
#p.add_tools(HoverTool(tooltips=IMG_TOOLTIPS))
inFile.text = fname
outDir.text = output_dir
else:
print('Cancelled. To continue please set output directory.{:<100}'.format(' '),end="\r")
# resetting sources for new data or new filters/contours
def remove_plot():
source.data["num_contours"]=[0]
contours_found.text = 'Droplets Detected: 0'
source_contours.data = dict(xs=[], ys=[])
source_label.data = dict(x=[], y=[], label=[])
# apply threshold filter and display binary image
def apply_filter():
if source.data['input_dir'][0] == '':
print('No image loaded! Load image first.{:<100}'.format(' '),end="\r")
else:
img = np.squeeze(source.data['image_orig'])
# remove contours if present
if source.data["num_contours"]!=[0]:
remove_plot()
if radio_button_gp.active == 1:
thresh = filters.threshold_otsu(img)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
elif radio_button_gp.active == 2:
thresh = filters.threshold_isodata(img)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
elif radio_button_gp.active == 3:
thresh = filters.threshold_mean(img)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
elif radio_button_gp.active == 4:
thresh = filters.threshold_li(img)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
elif radio_button_gp.active == 5:
thresh = filters.threshold_yen(img)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
elif radio_button_gp.active == 6:
off = offset_spinner.value
block_size = block_spinner.value
thresh = filters.threshold_local(img,block_size,offset=off)
binary = img > thresh
bin_img = binary*255
source.data["bin_img"] = [bin_img]
else:
bin_img = img
source.data['image'] = [bin_img]
# image functions for adjusting the binary image
def close_img():
if source.data["num_contours"]!=[0]:
remove_plot()
if radio_button_gp.active == 0:
print("Please Select Filter for Threshold{:<100}".format(' '),end="\r")
else:
source.data["image"] = source.data["bin_img"]
img = np.squeeze(source.data['bin_img'])
closed_img = binary_closing(255-img)*255
source.data['image'] = [255-closed_img]
source.data['bin_img'] = [255-closed_img]
def dilate_img():
if source.data["num_contours"]!=[0]:
remove_plot()
if radio_button_gp.active == 0:
print("Please Select Filter for Threshold{:<100}".format(' '),end="\r")
else:
img = np.squeeze(source.data['bin_img'])
dilated_img = binary_dilation(255-img)*255
source.data['image'] = [255-dilated_img]
source.data['bin_img'] = [255-dilated_img]
def erode_img():
if source.data["num_contours"]!=[0]:
remove_plot()
if radio_button_gp.active == 0:
print("Please Select Filter for Threshold{:<100}".format(' '),end="\r")
else:
img = np.squeeze(source.data['bin_img'])
eroded_img = binary_erosion(255-img)*255
source.data['image'] = [255-eroded_img]
source.data['bin_img'] = [255-eroded_img]
# the function for identifying closed contours in the image
def find_contours(level=0.8):
min_drop_size = contour_rng_slider.value[0]
max_drop_size = contour_rng_slider.value[1]
min_dim = 20
max_dim = 200
if radio_button_gp.active == 0:
print("Please Select Filter for Threshold{:<100}".format(' '),end="\r")
elif source.data['input_dir'][0] == '':
print('No image loaded! Load image first.{:<100}'.format(' '),end="\r")
else:
img = np.squeeze(source.data['bin_img'])
h,w = np.shape(img)
contours = measure.find_contours(img, level)
length_cnt_x = [cnt[:,1] for cnt in contours if np.shape(cnt)[0] < max_drop_size
and np.shape(cnt)[0] > min_drop_size]
length_cnt_y = [cnt[:,0] for cnt in contours if np.shape(cnt)[0] < max_drop_size
and np.shape(cnt)[0] > min_drop_size]
matched_cnt_x = []
matched_cnt_y = []
roi_coords = []
label_text = []
label_y = np.array([])
count=0
for i in range(len(length_cnt_x)):
cnt_x = length_cnt_x[i]
cnt_y = length_cnt_y[i]
width = np.max(cnt_x)-np.min(cnt_x)
height = np.max(cnt_y)-np.min(cnt_y)
if width>min_dim and width<max_dim and height>min_dim and height<max_dim:
matched_cnt_x.append(cnt_x)
matched_cnt_y.append(cnt_y)
roi_coords.append([round(width),round(height),round(np.min(cnt_x)),round(h-np.max(cnt_y))])
label_text.append(str(int(count)+1))
label_y = np.append(label_y,np.max(cnt_y))
count+=1
curr_contours = p.select_one({'name':'overlay'})
if curr_contours:
p.renderers.remove(curr_contours)
#if source.data["num_contours"]==[0]:
#remove_plot()
#p.image(source=source, image='image_orig', x=0, y=0, dw=w, dh=h, color_mapper=cmap, name='overlay',level='underlay')
source.data["image"] = source.data["image_orig"]
source.data["num_contours"] = [count]
#source.data["cnt_x"] = [matched_cnt_x]
#source.data["cnt_y"] = [matched_cnt_y]
source.data["roi_coords"] = [roi_coords]
source_contours.data = dict(xs=matched_cnt_x, ys=matched_cnt_y)
p.multi_line(xs='xs',ys='ys',source=source_contours, color=(255,127,14),line_width=2, name="contours",level='glyph')
if len(np.array(roi_coords).shape)<2:
if len(np.array(roi_coords)) <4:
print('No contours found. Try adjusting parameters or filter for thresholding.{:<100}'.format(' '),end="\r")
source_label.data = dict(x=[],y=[],label=[])
else:
source_label.data = dict(x=np.array(roi_coords)[2], y=label_y, label=label_text)
else:
source_label.data = dict(x=np.array(roi_coords)[:,2], y=label_y, label=label_text)
contours_found.text = 'Droplets Detected: '+str(len(matched_cnt_x))
# write the contours and parameters to files
def export_ROIs():
if source.data["num_contours"]==[0]:
print("No Contours Found! Find contours first.{:<100}".format(' '),end="\r")
else:
hdr = 'threshold filter,contour min,contour max'
thresh_filter = radio_button_gp.active
cnt_min = contour_rng_slider.value[0]
cnt_max = contour_rng_slider.value[1]
params = [thresh_filter,cnt_min,cnt_max]
if radio_button_gp.active == 6:
off = offset_spinner.value
block_size = block_spinner.value
hdr = hdr + ',local offset,local block size'
params.append(off,block_size)
params_fname = 'ContourParams.csv'
params_out = os.path.join(source.data['output_dir'][0],params_fname)
overwrite = 'no'
if os.path.exists(params_out):
overwrite = overwrite_file()
if overwrite == 'yes' or not os.path.exists(params_out):
np.savetxt(params_out,np.array([params]),delimiter=',',header=hdr,comments='')
roi_coords = np.array(source.data["roi_coords"][0])
out_fname = 'ROI_coords.csv'
out_fullpath = os.path.join(source.data['output_dir'][0],out_fname)
if overwrite == 'yes' or not os.path.exists(out_fullpath):
hdr = 'width,height,x,y'
np.savetxt(out_fullpath,roi_coords,delimiter=',',header=hdr,comments='')
print('Successfully saved ROIs coordinates as: '+out_fullpath,end='\r')
source.data['roi_coords'] = [roi_coords]
# function for loading previously made files or error handling for going out of order
def check_ROI_files():
coords_file = os.path.join(source.data["output_dir"][0],'ROI_coords.csv')
n_cnt_curr = source.data["num_contours"][0]
roi_coords_curr = source.data["roi_coords"][0]
if os.path.exists(coords_file):
df_tmp=pd.read_csv(coords_file, sep=',')
roi_coords = np.array(df_tmp.values)
n_cnt = len(roi_coords)
if n_cnt_curr != n_cnt or not np.array_equal(roi_coords_curr,roi_coords):
print('Current ROIs are different from saved ROI file! ROIs from saved file will be used instead and plot updated.',end="\r")
source.data["num_contours"] = [n_cnt]
source.data["roi_coords"] = [roi_coords]
params_file = os.path.join(source.data['output_dir'][0],'ContourParams.csv')
params_df = pd.read_csv(params_file)
thresh_ind = params_df["threshold filter"].values[0]
radio_button_gp.active = int(thresh_ind)
if thresh_ind == 6:
offset_spinner.value = int(params_df["local offset"].values[0])
block_spinner.value = int(params_df["local block size"].values[0])
contour_rng_slider.value = tuple([int(params_df["contour min"].values[0]),int(params_df["contour max"].values[0])])
find_contours()
else:
print("ROI files not found! Check save directory or export ROIs.{:<100}".format(' '),end="\r")
# use FFMPEG to crop out regions from original mp4 and save to file
def create_ROI_movies():
if source.data['input_dir'][0] == '':
print('No image loaded! Load image first.{:<100}'.format(' '),end="\r")
else:
check_ROI_files()
side = 100 # for square ROIs, replace first two crop parameters with side & uncomment
padding = 20
roi_coords_file = os.path.join(source.data['output_dir'][0],'ROI_coords.csv')
if source.data["num_contours"]==[0]:
print("No contours found! Find contours first.{:<100}".format(' '),end="\r")
elif not os.path.exists(roi_coords_file):
print("ROI file does not exist! Check save directory or export ROIs.{:<100}".format(' '),end="\r")
else:
print('Creating Movies...{:<100}'.format(' '),end="\r")
pbar = tqdm(total=source.data["num_contours"][0])
for i in range(source.data["num_contours"][0]):
roi_coords = np.array(source.data["roi_coords"][0])
inPath = os.path.join(source.data['input_dir'][0],source.data['img_name'][0])
#out_fname = source.data['filename'][0].split('.')[0] +'_ROI'+str(i+1)+'.mp4'
out_fname = 'ROI'+str(i+1)+'.mp4'
outPath = os.path.join(source.data['output_dir'][0],out_fname)
#command = f"ffmpeg -i \'{(inPath)}\' -vf \"crop={(roi_coords[i,0]+padding*2)}:{(roi_coords[i,1]+padding*2)}:{(roi_coords[i,2]-padding)}:{(roi_coords[i,3]+padding)}\" -y \'{(outPath)}\'"
command = f"ffmpeg -i \'{(inPath)}\' -vf \"crop={side}:{side}:{(roi_coords[i,2]-padding)}:{(roi_coords[i,3]-padding)}\" -y \'{(outPath)}\'"
overwrite = 'no'
if os.path.exists(outPath):
overwrite = overwrite_file()
if overwrite == 'yes' or not os.path.exists(outPath):
saved = subprocess.check_call(command,shell=True)
if saved != 0:
print('An error occurred while creating movies! Check terminal window.{:<100}'.format(' '),end="\r")
pbar.update()
# change the display range on images from slider values
def update_image():
cmap.low = display_range_slider.value[0]
cmap.high = display_range_slider.value[1]
# create statistics files for each mp4 region specific file
def process_ROIs():
if source.data['input_dir'][0] == '':
print('No image loaded! Load image first.{:<100}'.format(' '),end="\r")
else:
check_ROI_files()
hdr = 'roi,time,area,mean,variance,min,max,median,skewness,kurtosis,rawDensity,COMx,COMy'
cols = hdr.split(',')
all_stats = np.zeros((0,13))
n_cnt = source.data["num_contours"][0]
if n_cnt == 0:
print("No contours found! Find contours first.{:<100}".format(' '),end="\r")
for i in range(n_cnt):
#in_fname = source.data['filename'][0].split('.')[0] +'_ROI'+str(i+1)+'.mp4'
in_fname = 'ROI'+str(i+1)+'.mp4'
inPath = os.path.join(source.data['output_dir'][0],in_fname)
#out_fname = source.data['filename'][0].split('.')[0] +'_ROI'+str(i+1)+'_stats.csv'
out_fname = 'stats_ROI'+str(i+1)+'.csv'
outPath = os.path.join(source.data['output_dir'][0],out_fname)
if not os.path.exists(inPath):
print('ROI movie not found! Create ROI movie first.{:<100}'.format(' '),end="\r")
break
vidcap = cv2.VideoCapture(inPath)
last_frame = int(vidcap.get(cv2.CAP_PROP_FRAME_COUNT))
if i==0:
pbar = tqdm(total=last_frame*n_cnt)
success,frame = vidcap.read()
img_tmp,_,__ = cv2.split(frame)
h,w = np.shape(img_tmp)
img = np.zeros((last_frame,h,w))
img_stats = np.zeros((last_frame,13))
stats = describe(img_tmp,axis=None)
median = np.median(img_tmp)
density = np.sum(img_tmp)
cx, cy = center_of_mass(img_tmp)
img_stats[0,0:13] = [i,0,stats.nobs,stats.mean,stats.variance,
stats.minmax[0],stats.minmax[1],median,stats.skewness,
stats.kurtosis,density,cx,cy]
img[0,:,:] = np.flipud(img_tmp)
pbar.update()
overwrite = 'no'
if os.path.exists(outPath):
overwrite = overwrite_file()
if overwrite=='no':
pbar.update(last_frame-1)
if overwrite == 'yes' or not os.path.exists(outPath):
for j in range(1,last_frame):
vidcap.set(1, j)
success,frame = vidcap.read()
img_tmp,_,__ = cv2.split(frame)
stats = describe(img_tmp,axis=None)
t = j*5/60
density = np.sum(img_tmp)
cx, cy = center_of_mass(img_tmp)
median = np.median(img_tmp)
img_stats[j,0:13] = [i,t,stats.nobs,stats.mean,stats.variance,
stats.minmax[0],stats.minmax[1],median,stats.skewness,
stats.kurtosis,density,cx,cy]
img[j,:,:] = np.flipud(img_tmp)
pbar.update(1)
all_stats = np.append(all_stats,img_stats,axis=0)
np.savetxt(outPath,img_stats,delimiter=',',header=hdr,comments='')
if i==(n_cnt-1):
df = pd.DataFrame(all_stats,columns=cols)
group = df.groupby('roi')
for i in range(len(group)):
sources_stats[i] = ColumnDataSource(group.get_group(i))
# load statistics CSVs and first ROI mp4 files and display in plots
def load_ROI_files():
if source.data['input_dir'][0] == '':
print('No image loaded! Load image first.{:<100}'.format(' '),end="\r")
else:
check_ROI_files()
n_cnt = source.data["num_contours"][0]
basepath = os.path.join(source.data["output_dir"][0],'stats')
all_files = [basepath+'_ROI'+str(i+1)+'.csv' for i in range(n_cnt)]
files_exist = [os.path.exists(f) for f in all_files]
if all(files_exist) and n_cnt != 0:
df = pd.concat((pd.read_csv(f) for f in all_files), ignore_index=True)
group = df.groupby('roi')
OPTIONS = []
LABELS = []
pbar = tqdm(total=len(stats)*len(group))
j=0
colors_ordered = list(Category20[20])
idx_reorder = np.append(np.linspace(0,18,10),np.linspace(1,19,10))
idx = idx_reorder.astype(int)
colors = [colors_ordered[i] for i in idx]
for roi, df_roi in group:
sources_stats[roi] = ColumnDataSource(df_roi)
OPTIONS.append([str(int(roi)+1),'ROI '+(str(int(roi)+1))])
LABELS.append('ROI '+str(int(roi)+1))
color = colors[j]
j+=1
if j>=20:
j=0
for i in range(3,len(df.columns)):
name = 'ROI '+str(int(roi)+1)
plot_check = p_stats[i-3].select_one({'name':name})
if not plot_check:
p_stats[i-3].line(x='time',y=str(df.columns[i]),source=sources_stats[roi],
name=name,visible=False,line_color=color)
p_stats[i-3].xaxis.axis_label = "Time [h]"
p_stats[i-3].yaxis.axis_label = str(df.columns[i])
p_stats[i-3].add_tools(HoverTool(tooltips=TOOLTIPS))
p_stats[i-3].toolbar_location = "left"
pbar.update(1)
ROI_multi_select.options = OPTIONS
ROI_multi_select.value = ["1"]
ROI_movie_radio_group.labels = LABELS
ROI_movie_radio_group.active = 0
else:
print('Not enough files! Check save directory or calculate new stats.{:<100}'.format(' '),end="\r")
# show/hide curves from selected/deselected labels for ROIs in statistics plots
def update_ROI_plots():
n_cnt = source.data["num_contours"][0]
pbar = tqdm(total=len(stats)*n_cnt)
for j in range(n_cnt):
for i in range(len(stats)):
name = 'ROI '+str(int(j)+1)
glyph = p_stats[i].select_one({'name': name})
if str(j+1) in ROI_multi_select.value:
glyph.visible = True
else:
glyph.visible = False
pbar.update(1)
# load and display the selected ROI's mp4
def load_ROI_movie():
idx = ROI_movie_radio_group.active
in_fname = 'ROI'+str(idx+1)+'.mp4'
inPath = os.path.join(source.data['output_dir'][0],in_fname)
if not os.path.exists(inPath):
print('ROI movie not found! Check save directory or create ROI movie.',end="\r")
else:
old_plot = p_ROI.select_one({'name': sourceROI.data['img_name'][0]})
if old_plot:
p_ROI.renderers.remove(old_plot)
vidcap = cv2.VideoCapture(inPath)
last_frame = int(vidcap.get(cv2.CAP_PROP_FRAME_COUNT))
ROI_movie_slider.end = (last_frame-1)*5/60
ROI_movie_slider.value = 0
vidcap.set(1, 0)
success,frame = vidcap.read()
img_tmp,_,__ = cv2.split(frame)
h,w = np.shape(img_tmp)
img = np.flipud(img_tmp)
name = 'ROI'+str(idx+1)
sourceROI.data = dict(image=[img],x=[0], y=[0], dw=[w], dh=[h],
img_name=[name])
p_ROI.image(source=sourceROI, image='image', x='x', y='y',
dw='dw', dh='dh', color_mapper=cmap, name='img_name')
# change the displayed frame from slider movement
def update_ROI_movie():
frame_idx = round(ROI_movie_slider.value*60/5)
in_fname = sourceROI.data['img_name'][0]+'.mp4'
inPath = os.path.join(source.data['output_dir'][0],in_fname)
vidcap = cv2.VideoCapture(inPath)
vidcap.set(1, frame_idx)
success,frame = vidcap.read()
img_tmp,_,__ = cv2.split(frame)
img = np.flipud(img_tmp)
sourceROI.data['image'] = [img]
# the following 2 functions are used to animate the mp4
def update_ROI_slider():
time = ROI_movie_slider.value + 5/60
end = ROI_movie_slider.end
if time > end:
animate_ROI_movie()
else:
ROI_movie_slider.value = time
return callback_id
def animate_ROI_movie():
global callback_id
if ROI_movie_play_button.label == '► Play':
ROI_movie_play_button.label = '❚❚ Pause'
callback_id = curdoc().add_periodic_callback(update_ROI_slider, 10)
else:
ROI_movie_play_button.label = '► Play'
curdoc().remove_periodic_callback(callback_id)
return callback_id
### Application Content ###
# main plot for segmentation and contour finding
cmap = LinearColorMapper(palette="Greys256", low=0, high=255)
TOOLS = "pan,wheel_zoom,box_zoom,reset,save,box_select,lasso_select"
IMG_TOOLTIPS = [('name', "@img_name"),("x", "$x"),("y", "$y"),("value", "@image")]
source = ColumnDataSource(data=dict(image=[0],bin_img=[0],image_orig=[0],
x=[0], y=[0], dw=[0], dh=[0], num_contours=[0], roi_coords=[0],
input_dir=[''],output_dir=[''],img_name=['']))
source_label = ColumnDataSource(data=dict(x=[0], y=[0], label=['']))
source_contours = ColumnDataSource(data=dict(xs=[0], ys=[0]))
roi_labels = LabelSet(x='x', y='y', text='label',source=source_label,
level='annotation',text_color='white',text_font_size='12pt')
# create a new plot and add a renderer
p = figure(tools=TOOLS, toolbar_location=("right"))
p.add_layout(roi_labels)
p.x_range.range_padding = p.y_range.range_padding = 0
# turn off gridlines
p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None
p.axis.visible = False
# ROI plots
sourceROI = ColumnDataSource(data=dict(image=[0],
x=[0], y=[0], dw=[0], dh=[0], img_name=[0]))
sources_stats = {}
TOOLTIPS = [('name','$name'),('time', '@time'),('stat', "$y")]
stats = np.array(['mean','var','min','max','median','skew','kurt','rawDensity','COMx','COMy'])
p_stats = []
tabs = []
for i in range(len(stats)):
p_stats.append(figure(tools=TOOLS, plot_height=300, plot_width=600))
p_stats[i].x_range.range_padding = p_stats[i].y_range.range_padding = 0
tabs.append(Panel(child=p_stats[i], title=stats[i]))
# create a new plot and add a renderer
p_ROI = figure(tools=TOOLS, toolbar_location=("right"), plot_height=300, plot_width=300)
p_ROI.x_range.range_padding = p_ROI.y_range.range_padding = 0
# turn off gridlines
p_ROI.xgrid.grid_line_color = p_ROI.ygrid.grid_line_color = None
p_ROI.axis.visible = False
# Widgets - Buttons, Sliders, Text, Etc.
intro = Div(text="""<h2>Droplet Recognition and Analysis with Bokeh</h2>
This application is designed to help segment a grayscale image into
regions of interest (ROIs) and perform analysis on those regions.<br>
<h4>How to Use This Application:</h4>
<ol>
<li>Load in a grayscale mp4 file and choose a save directory.</li>
<li>Apply various filters for thresholding. Use <b>Close</b>, <b>Dilate</b>
and <b>Erode</b> buttons to adjust each binary image further.</li>
<li>Use <b>Find Contours</b> button to search the image for closed shape.
The <b>Contour Size Range</b> slider will change size of the perimeter to
be identified. You can apply new thresholds and repeat until satisfied with
the region selection. Total regions detected is displayed next to
the button.</li>
<li>When satisfied, use <b>Export ROIs</b> to write ROI locations and
contour finding parameters to file.</li>
<li><b>Create ROI Movies</b> to write mp4s of the selected regions.</li>
<li>Use <b>Calculate ROI Stats</b> to perform calculations on the
newly created mp4 files.</li>
<li>Finally, use <b>Load ROI Files</b> to load in the data that you just
created and view the plots. The statistics plots can be overlaid by
selecting multiple labels. Individual ROI mp4s can be animated or you can
use the slider to move through the frames.</li>
</ol>
Note: messages and progress bars are displayed below the GUI.""",
style={'font-size':'10pt'},width=1000)
file_button = Button(label="Choose File",button_type="primary")
file_button.on_click(update_filename)
inFile = PreText(text='Input File:\n'+source.data["img_name"][0], background=(255,255,255,0.5), width=500)
filter_LABELS = ["Original","OTSU", "Isodata", "Mean", "Li","Yen","Local"]
radio_button_gp = RadioButtonGroup(labels=filter_LABELS, active=0, width=600)
radio_button_gp.on_change('active', lambda attr, old, new: apply_filter())
offset_spinner = Spinner(low=0, high=500, value=1, step=1, width=100, title="Local: Offset",
background=(255,255,255,0.5))
offset_spinner.on_change('value', lambda attr, old, new: apply_filter())
block_spinner = Spinner(low=1, high=101, value=25, step=2, width=100, title="Local: Block Size",
background=(255,255,255,0.5))
block_spinner.on_change('value', lambda attr, old, new: apply_filter())
closing_button = Button(label="Close",button_type="default", width=100)
closing_button.on_click(close_img)
dilation_button = Button(label="Dilate",button_type="default", width=100)
dilation_button.on_click(dilate_img)
erosion_button = Button(label="Erode",button_type="default", width=100)
erosion_button.on_click(erode_img)
contour_rng_slider = RangeSlider(start=10, end=500, value=(200,350), step=1, width=300,
title="Contour Size Range", background=(255,255,255,0.5), bar_color='gray')
contour_button = Button(label="Find Contours", button_type="success")
contour_button.on_click(find_contours)
contours_found = PreText(text='Droplets Detected: '+str(source.data["num_contours"][0]), background=(255,255,255,0.5))
exportROIs_button = Button(label="Export ROIs", button_type="success", width=200)
exportROIs_button.on_click(export_ROIs)
changeDir_button = Button(label="Change Directory",button_type="primary", width=150)
changeDir_button.on_click(change_directory)
outDir = PreText(text='Save Directory:\n'+source.data["output_dir"][0], background=(255,255,255,0.5), width=500)
create_ROIs_button = Button(label="Create ROI Movies",button_type="success", width=200)
create_ROIs_button.on_click(create_ROI_movies)
process_ROIs_button = Button(label="Calculate ROI Stats",button_type="success")
process_ROIs_button.on_click(process_ROIs)
display_rng_text = figure(title="Display Range", title_location="left",
width=40, height=300, toolbar_location=None, min_border=0,
outline_line_color=None)
display_rng_text.title.align="center"
display_rng_text.title.text_font_size = '10pt'
display_rng_text.x_range.range_padding = display_rng_text.y_range.range_padding = 0
display_range_slider = RangeSlider(start=0, end=255, value=(0,255), step=1,
orientation='vertical', direction='rtl',
bar_color='gray', width=40, height=300, tooltips=True)
display_range_slider.on_change('value', lambda attr, old, new: update_image())
load_ROIfiles_button = Button(label="Load ROI Files",button_type="primary")
load_ROIfiles_button.on_click(load_ROI_files)
ROI_multi_select = MultiSelect(value=[], width=100, height=300)
ROI_multi_select.on_change('value', lambda attr, old, new: update_ROI_plots())
ROI_movie_radio_group = RadioGroup(labels=[],width=60)
ROI_movie_radio_group.on_change('active', lambda attr, old, new: load_ROI_movie())
ROI_movie_slider = Slider(start=0,end=100,value=0,step=5/60,title="Time [h]", width=280)
ROI_movie_slider.on_change('value', lambda attr, old, new: update_ROI_movie())
callback_id = None
ROI_movie_play_button = Button(label='► Play',width=50)
ROI_movie_play_button.on_click(animate_ROI_movie)
# initialize some data without having to choose file
# fname = os.path.join(os.getcwd(),'data','Droplets.mp4')
# load_data(filename=fname)
### Layout & Initialize application ###
ROI_layout = layout([
[ROI_movie_radio_group, p_ROI],
[ROI_movie_slider,ROI_movie_play_button]
])
app = layout(children=[
[intro],
[file_button,inFile],
[radio_button_gp, offset_spinner, block_spinner],
[closing_button, dilation_button, erosion_button],
[contour_rng_slider, contour_button, contours_found],
[exportROIs_button, outDir, changeDir_button],
[create_ROIs_button, process_ROIs_button],
[display_rng_text, display_range_slider, p],
[load_ROIfiles_button],
[ROI_layout, ROI_multi_select, Tabs(tabs=tabs)]
])
doc.add_root(app)
df = load_data()
##########
########## Preprocess data
##########
source = preprocess_data(df)
##########
########## Create checkerbox filter
##########
years_list = sorted(df['CalendarYearIssued'].astype(str).unique().tolist())
year_selection = MultiSelect(title='Year'
, value=[str(i) for i,j in enumerate(years_list)]
, options = [(str(i),j) for i,j in enumerate(years_list)])
##########
########## Create select all button
##########
def update_selectall():
year_selection.value = [x[0] for x in year_selection.options]
select_all = Button(label='Select All')
select_all.on_click(update_selectall)
##########
########## Create run button
def create_world_cases_time_series_tab():
## Data Sources
source_df, source_CDS = get_country_cases_vs_time()
## Line Plots
line_figure = figure(
x_axis_type='datetime',
y_axis_type='log',
title='World Confirmed Cases by Region',
x_axis_label='Date',
y_axis_label='Number of Confirmed Cases (Logarithmic Scale)',
active_scroll='wheel_zoom')
starting_regions = ['China', 'US', 'Italy']
excluded_columns_set = {'index', 'date'}
doubling_lines_props = {
'alpha': 0.6,
'muted_alpha': 0.2,
'line_width': 3,
'source': source_CDS,
'x': 'date',
'visible': True
}
for number, text, color in zip([4, 7, 14], ['four', 'seven', 'fourteen'],
gray(6)[2:5]):
column_name = f'{text}_day_doubling'
excluded_columns_set.add(column_name)
source_CDS.data[column_name] = 2**(
np.arange(len(source_CDS.data['index'])) / number)
line_figure.line(y=column_name,
legend_label=f'{number}-day Doubling Time',
line_color=color,
name=column_name,
**doubling_lines_props)
line_params = {
'x': 'date',
'source': source_CDS,
'line_width': 4,
'alpha': 0.6
}
lines = {
key: line_figure.line(y=key, name=key, line_color=color, **line_params)
for key, color in zip(starting_regions, viridis(len(starting_regions)))
}
line_figure.legend.location = 'top_left'
line_figure.legend.click_policy = 'hide'
hover_tool = HoverTool(
tooltips=[('Date', '@date{%F}'), ('Region', '$name'),
('Number of Cases', '@$name{0,0}')],
formatters={
'@date': 'datetime',
},
renderers=[*line_figure.renderers],
# mode='vline'
)
line_figure.add_tools(hover_tool)
## Region Selector
labels = [
key for key in source_CDS.data.keys()
if key not in excluded_columns_set
]
def region_select_callback(attr, old, new):
new_lines = set(new) - set(old)
old_lines = set(old) - set(new)
for key in old_lines:
lines[key].visible = False
for key in new_lines:
if key in lines.keys():
lines[key].visible = True
else:
lines[key] = line_figure.line(
y=key,
name=key,
line_color=np.random.choice(Viridis256),
**line_params)
hover_tool.renderers = [*hover_tool.renderers, lines[key]]
region_select = MultiSelect(title='Select Regions to Show',
value=starting_regions,
options=labels,
sizing_mode='stretch_height')
region_select.on_change('value', region_select_callback)
## Create Layout
child = row([
column([line_figure]),
column([region_select]),
])
return Panel(child=child, title='World Cases Time Series')
bottom_row.children[0] = cost_plot
# Parse arguments and connect to db
parser = create_parser()
args = sys.argv[1:]
parsed_args = parser.parse_args(args=args)
conn = connect_to_database(db_path=parsed_args.database)
# Set Up Data
data = DataProvider(conn)
# Set up selection widgets
scenario_select = MultiSelect(
title="Select Scenario(s):",
value=data.scenario_options,
# width=600,
options=data.scenario_options)
period_select = MultiSelect(title="Select Period(s):",
value=data.period_options,
options=data.period_options)
stage_select = Select(title="Select Stage:",
value=data.stage_options[0],
options=data.stage_options)
zone_select = Select(title="Select Load Zone:",
value=data.zone_options[0],
options=data.zone_options)
capacity_select = Select(title="Select Capacity Metric:",
value=data.cap_options[2],
options=data.cap_options)
return (settings_files_select_options)
make_cases_div = Div(text="<h2>Make case(s)</h2>", sizing_mode="stretch_width")
# Settings file picker
make_cases_select_div = Div(text="Select one settings file.",
css_classes=['item-description'],
sizing_mode='stretch_width')
# List for picking a file
get_settings_files()
settings_files_select_options = retrieve_select_options(settings_file_names)
make_cases_select = MultiSelect(value=[],
options=settings_files_select_options)
# Make cases button
make_cases_button = Button(label="Make case(s)",
button_type='primary',
disabled=False)
refresh_button = Button(label="Refresh files",
button_type='default',
disabled=False)
make_cases_output = Div(text="Click the button above to create and build " +
"the cases specified in the chosen settings file.",
css_classes=["item-description"],
style={
'overflow': 'auto',
'width': '100%',
class Selector:
def __init__(
self,
name="Specials",
descr="Choose one",
kind="specials",
css_classes=[],
entries={},
default="",
title=None,
none_allowed=False,
):
self.name = name
self.descr = descr
self.entries = entries
self.kind = kind
self.css_classes = css_classes
options = sorted(entries.keys())
if none_allowed:
options = ["None"] + options
if title is None:
title = "."
css_classes = ["deli-selector", "hide-title"]
else:
css_classes = ["deli-selector"]
self.widget = MultiSelect(
options=options,
value=[default],
# height=150,
size=8,
name="deli-selector",
title=title,
css_classes=css_classes,
)
# HACK: force MultiSelect to only have 1 value selected
def multi_select_hack(attr, old, new):
if len(new) > 1:
self.widget.value = old
self.widget.on_change("value", multi_select_hack)
@property
def value(self):
# HACK: This is because we are useing MultiSelect instead of Select
return self.widget.value[0]
def layout(self, additional_widgets=[], width=None):
title = Div(
text="""<h2>{0}</h2><h3>{1}</h3>""".format(self.name, self.descr),
css_classes=["controls"],
)
footer = Div(
text="""<a href="#">About the {0}</a>""".format(self.kind),
css_classes=["controls", "controls-footer"],
)
if width is None:
width = 160 * (1 + len(additional_widgets))
return column(
title,
row(
self.widget,
*additional_widgets,
width=width,
css_classes=["controls"],
),
footer,
css_classes=self.css_classes,
)
def build_graph_plot(G, title=""):
""" Return a Bokeh plot of the given networkx graph
Parameters
----------
G: :obj:`networkx.Graph`
Networkx graph instance to be plotted.
title: str
Title of the final plot
Returns
-------
:obj:`bokeh.models.plot`
Bokeh plot of the graph.
"""
plot = Plot(plot_width=600,
plot_height=450,
x_range=Range1d(-1.1, 1.1),
y_range=Range1d(-1.1, 1.1))
plot.title.text = title
node_attrs = {}
for node in G.nodes(data=True):
node_color = Spectral4[node[1]['n']]
node_attrs[node[0]] = node_color
nx.set_node_attributes(G, node_attrs, "node_color")
node_hover_tool = HoverTool(tooltips=[("Label", "@label"), ("n", "@n")])
wheelZoom = WheelZoomTool()
plot.add_tools(node_hover_tool, PanTool(), wheelZoom, ResetTool())
plot.toolbar.active_scroll = wheelZoom
graph_renderer = from_networkx(G,
nx.spring_layout,
k=0.3,
iterations=200,
scale=1,
center=(0, 0))
graph_renderer.node_renderer.glyph = Circle(size=15,
fill_color="node_color")
graph_renderer.edge_renderer.glyph = MultiLine(line_alpha=0.8,
line_width=1)
plot.renderers.append(graph_renderer)
selectCallback = CustomJS(args=dict(graph_renderer=graph_renderer),
code="""
let new_data_nodes = Object.assign({},graph_renderer.node_renderer.data_source.data);
new_data_nodes['node_color'] = {};
let colors = ['#2b83ba','#ABDDA4','#fdae61'];
let ns = cb_obj.value.reduce((acc,v)=>{
if(v=='fullGraph'){
acc.push(0);
acc.push(1);
acc.push(2);
}
if(v=='n0')acc.push(0);
if(v=='n1')acc.push(1);
if(v=='n2')acc.push(2);
return acc;
},[])
Object.keys(graph_renderer.node_renderer.data_source.data['node_color']).map((n,i)=>{
new_data_nodes['node_color'][i]='transparent';
})
ns.map(n=>{
Object.keys(graph_renderer.node_renderer.data_source.data['node_color']).map((g,i)=>{
if(graph_renderer.node_renderer.data_source.data['n'][i]==n){
new_data_nodes['node_color'][i]=colors[n];
}
})
})
graph_renderer.node_renderer.data_source.data = new_data_nodes
""")
multi_select = MultiSelect(title="Option:",
options=[("fullGraph", "Full Graph"),
("n0", "Seed Nodes"), ("n1", "N1"),
("n2", "N2")])
multi_select.js_on_change('value', selectCallback)
return column(plot, multi_select)
BSURF_TEXT = PreText(text='', width=500)
EDOT_TEXT = PreText(text='', width=500)
status = PreText(text="""TODO:
1. Make P-Pdot plot more useful.
2. Add user comments.
3. Discuss classification tags.
4. Feedback.
5. Add histograms.
""",
width=500)
CATALOGUE = Toggle(label="Toggle catalogue", width=300, button_type="success")
UPDATE = Button(label="Update", width=300)
PROFILE = MultiSelect(title="Profile tags",
options=tags["PROFILE"],
value=["Unclassified"],
height=200)
POL = MultiSelect(title="Polarization tags",
options=tags["POLARIZATION"],
value=["Unclassified"],
height=200)
FREQ = MultiSelect(title="Frequency tags",
options=tags["FREQUENCY"],
value=["Unclassified"],
height=200)
TIME = MultiSelect(title="Time tags",
options=tags["TIME"],
value=["Unclassified"],
height=200)
OBS = MultiSelect(title="Observation tags",
options=tags["OBSERVATION"],
'y': y,
'label': label
}) #create a dataframe for future use
source = ColumnDataSource(data=dict(x=x, y=y, label=label))
plot_figure = figure(title='Multi-Select',
height=450,
width=600,
tools="save,reset",
toolbar_location="below")
plot_figure.scatter('x', 'y', color='label', source=source, size=10)
multi_select = MultiSelect(title="Filter Plot by color:",
value=["Red", "Orange"],
options=[("Red", "Red"), ("Orange", "Orange")])
def multiselect_click(attr, old, new):
active_mselect = multi_select.value ##Getting multi-select value
selected_df = df[df['label'].isin(
active_mselect)] #filter the dataframe with value in multi-select
source.data = dict(x=selected_df.x,
y=selected_df.y,
label=selected_df.label)
multi_select.on_change('value', multiselect_click)
def do_layout(self):
"""
generates the overall layout by creating all the widgets, buttons etc and arranges
them in rows and columns
:return: None
"""
self.source = self.generate_source()
tab_plot = self.generate_plot(self.source)
multi_select = MultiSelect(title="Option (Multiselect Ctrl+Click):",
value=self.active_country_list,
options=countries,
height=500)
multi_select.on_change('value', self.update_data)
tab_plot.on_change('active', self.update_tab)
radio_button_group_per_capita = RadioButtonGroup(
labels=["Total Cases", "Cases per Million"],
active=0 if not self.active_per_capita else 1)
radio_button_group_per_capita.on_click(self.update_capita)
radio_button_group_scale = RadioButtonGroup(
labels=[Scale.log.name.title(),
Scale.linear.name.title()],
active=self.active_y_axis_type.value)
radio_button_group_scale.on_click(self.update_scale_button)
radio_button_group_df = RadioButtonGroup(labels=[
Prefix.confirmed.name.title(),
Prefix.deaths.name.title(),
Prefix.recovered.name.title(),
],
active=int(
self.active_case_type))
radio_button_group_df.on_click(self.update_data_frame)
refresh_button = Button(label="Refresh Data",
button_type="default",
width=150)
refresh_button.on_click(load_data_frames)
export_button = Button(label="Export Url",
button_type="default",
width=150)
export_button.on_click(self.export_url)
slider = Slider(start=1,
end=30,
value=self.active_window_size,
step=1,
title="Window Size for rolling average")
slider.on_change('value', self.update_window_size)
radio_button_average = RadioButtonGroup(
labels=[Average.mean.name.title(),
Average.median.name.title()],
active=self.active_average)
radio_button_average.on_click(self.update_average_button)
plot_variables = [
self.active_plot_raw, self.active_plot_average,
self.active_plot_trend
]
plots_button_group = CheckboxButtonGroup(
labels=["Raw", "Averaged", "Trend"],
active=[i for i, x in enumerate(plot_variables) if x])
plots_button_group.on_click(self.update_shown_plots)
world_map = self.create_world_map()
link_div = Div(
name="URL",
text=
fr'Link <a target="_blank" href="{self.url}">Link to this Plot</a>.',
width=300,
height=10,
align='center')
footer = Div(
text=
"""Covid-19 Dashboard created by Andreas Weichslgartner in April 2020 with python, bokeh, pandas,
numpy, pyproj, and colorcet. Source Code can be found at
<a href="https://github.com/weichslgartner/covid_dashboard/">Github</a>.""",
width=1600,
height=10,
align='center')
self.generate_table_cumulative()
columns = [
TableColumn(field="name", title="Country"),
TableColumn(field="number_rolling",
title="daily avg",
formatter=NumberFormatter(format="0.")),
TableColumn(field="number_daily",
title="daily raw",
formatter=NumberFormatter(format="0."))
]
top_top_14_new_header = Div(text="Highest confirmed (daily)",
align='center')
top_top_14_new = DataTable(source=self.top_new_source,
name="Highest confirmed(daily)",
columns=columns,
width=300,
height=380)
self.generate_table_new()
columns = [
TableColumn(field="name", title="Country"),
TableColumn(field="number",
title="confirmed(cumulative)",
formatter=NumberFormatter(format="0."))
]
top_top_14_cum_header = Div(text="Highest confirmed (cumulative)",
align='center')
top_top_14_cum = DataTable(source=self.top_total_source,
name="Highest confirmed(cumulative)",
columns=columns,
width=300,
height=380)
self.layout = layout([
row(
column(tab_plot, world_map),
column(top_top_14_new_header, top_top_14_new,
top_top_14_cum_header, top_top_14_cum),
column(link_div, row(refresh_button, export_button),
radio_button_group_df, radio_button_group_per_capita,
plots_button_group, radio_button_group_scale, slider,
radio_button_average, multi_select),
),
row(footer)
])
curdoc().add_root(self.layout)
curdoc().title = "Bokeh Covid-19 Dashboard"
data_substantiated = pd.read_excel('NRC_allegation_stats.xlsx',
sheetname=2).set_index("Site").fillna(False)
data.fillna('')
data_substantiated.fillna('')
fruits = list(dataz['Site'])
years = [str(i) for i in data.columns.values]
palette = list(itertools.islice(palette, len(data.columns.values)))
# this creates [ ("Apples", "2015"), ("Apples", "2016"), ("Apples", "2017"), ("Pears", "2015), ... ]
years_widget = RangeSlider(start=2013,
end=2017,
value=(2015, 2017),
step=1,
title="Year[s]")
site_widget = MultiSelect(title="Site[s]",
value=["ARKANSAS 1 & 2"],
options=open(join(dirname(__file__),
'sites.txt')).read().split('\n'))
site_widget.on_change('value', lambda attr, old, new: update)
# this creates [ ("Apples", "2015"), ("Apples", "2016"), ("Apples", "2017"), ("Pears", "2015), ... ]
x = [(fruit, year) for fruit in fruits for year in years]
source = ColumnDataSource(data=dict(x=[], counts=[]))
source2 = ColumnDataSource(data=dict(x=[], counts=[]))
p = figure(x_range=data.index.values.tolist(),
plot_height=350,
title="Allegations Received from All Sources",
toolbar_location=None,
tools="")
p.vbar(x='x',
top='counts',
checkbox_button_group = CheckboxButtonGroup(labels=["Option 1", "Option 2", "Option 3"], active=[0, 1]) radio_button_group = RadioButtonGroup(labels=["Option 1", "Option 2", "Option 3"], active=0) checkbox_button_group_vertical = CheckboxButtonGroup(labels=["Option 1", "Option 2", "Option 3"], active=[0, 1], orientation="vertical") radio_button_group_vertical = RadioButtonGroup(labels=["Option 1", "Option 2", "Option 3"], active=0, orientation="vertical") text_input = TextInput(placeholder="Enter value ...") completions = ["aaa", "aab", "aac", "baa", "caa"] autocomplete_input = AutocompleteInput(placeholder="Enter value (auto-complete) ...", completions=completions) text_area = TextAreaInput(placeholder="Enter text ...", cols=20, rows=10, value="uuu") select = Select(options=["Option 1", "Option 2", "Option 3"]) multi_select = MultiSelect(options=["Option %d" % (i+1) for i in range(16)], size=6) multi_choice = MultiChoice(options=["Option %d" % (i+1) for i in range(16)]) slider = Slider(value=10, start=0, end=100, step=0.5) range_slider = RangeSlider(value=[10, 90], start=0, end=100, step=0.5) date_slider = DateSlider(value=date(2016, 1, 1), start=date(2015, 1, 1), end=date(2017, 12, 31)) date_range_slider = DateRangeSlider(value=(date(2016, 1, 1), date(2016, 12, 31)), start=date(2015, 1, 1), end=date(2017, 12, 31)) spinner = Spinner(value=100) color_picker = ColorPicker(color="red", title="Choose color:")
def __init__(self):
self.spe_file = None
self.full_sensor_data = None
self.selection_data = None
# setup widgets
self.directory_input = TextInput(placeholder='Directory',
value=os.path.join(
os.getcwd(), 'data'))
self.show_files_button = Button(label='Show Files',
button_type='primary')
self.file_view = MultiSelect(size=5)
self.open_file_button = Button(label='Open File',
button_type='warning')
self.update_selection_button = Button(label='Update Selection',
button_type='success')
self.selection_range = RangeSlider(start=0,
end=1,
value=(0, 1),
step=1,
title='Selected Rows')
# connect button callbacks
self.show_files_button.on_click(self.update_file_browser)
self.open_file_button.on_click(self.open_file_callback)
self.update_selection_button.on_click(self.update_selection)
self.selection_range.on_change('value', self.selection_range_callback)
# setup plots
self.full_sensor_image = figure(x_range=(0, 1),
y_range=(0, 1023),
tools='pan,box_zoom,wheel_zoom,reset',
plot_width=512,
plot_height=512)
self.full_sensor_image_label = Label(x=0.1,
y=0.1,
text='Source Data',
text_font_size='36pt',
text_color='#eeeeee')
self.full_sensor_image.add_tools(BoxSelectTool(dimensions='height'))
self.full_sensor_image.grid.grid_line_color = None
self.full_sensor_image.xaxis.major_tick_line_color = None
self.full_sensor_image.xaxis.minor_tick_line_color = None
self.full_sensor_image.yaxis.major_tick_line_color = None
self.full_sensor_image.yaxis.minor_tick_line_color = None
self.full_sensor_image.xaxis.major_label_text_font_size = '0pt'
self.full_sensor_image.yaxis.major_label_text_font_size = '0pt'
self.selection_lines_coords = ColumnDataSource(
data=dict(x=[[0, 1], [0, 1]], y=[[0, 0], [1, 1]]))
self.selection_image = figure(x_range=(0, 1),
y_range=(0, 1),
tools='wheel_zoom',
plot_width=1024,
plot_height=180)
self.selection_image_label = Label(x=0.1,
y=0.2,
text='Selection Region',
text_font_size='36pt',
text_color='#eeeeee')
self.selection_image.grid.grid_line_color = None
# build the layout
controls = [
self.directory_input, self.show_files_button, self.file_view,
self.open_file_button, self.selection_range,
self.update_selection_button
]
widgets = widgetbox(*controls, width=500)
self.layout = layout(
children=[[widgets],
[self.full_sensor_image, self.selection_image]],
sizing_mode='fixed')
# set defaults
self.initialize_ui()
class Trends:
"""Trends layout
"""
def __init__(self, palette=Purples[3]):
self.cases = LinePlot(ARIMA_CASES_TABLE)
self.cases.render_figure()
self.cases.title("Cumulative Cases by State")
self.cases.axis_label('Date', 'Cases')
self.cases.color_palette(palette)
LOG.debug('state cases')
self.deaths = LinePlot(ARIMA_DEATHS_TABLE)
self.deaths.render_figure()
self.deaths.title("Cumulative Deaths by State")
self.deaths.axis_label('Date', 'Deaths')
self.deaths.color_palette(palette)
LOG.debug('state deaths')
self.multiselect = None
self._add_multiselect()
self.multiselect.value = ['12', '34', '36']
LOG.debug('render default states')
def _add_multiselect(self):
self.multiselect = MultiSelect(title='States:', value=['01'],
options=self.cases.options)
self.multiselect.max_width = 170
self.multiselect.min_height = 500 - 47
self.multiselect.on_change('value', self._callback_cases)
self.multiselect.on_change('value', self._callback_deaths)
def _callback_cases(self, _attr, _old, new):
for _id, _ in list(self.multiselect.options):
if self.cases.actual[_id].visible:
self.cases.actual[_id].visible = False
self.cases.predict[_id].visible = False
self.cases.lower[_id].visible = False
self.cases.upper[_id].visible = False
self.cases.area[_id].visible = False
for _id in new:
if not self.cases.actual[_id].visible:
_slice = self.cases.data.loc[_id, :]
self.cases.source[_id].data = ColumnDataSource.from_df(data=_slice)
self.cases.actual[_id].visible = True
self.cases.predict[_id].visible = True
self.cases.lower[_id].visible = True
self.cases.upper[_id].visible = True
self.cases.area[_id].visible = True
def _callback_deaths(self, _attr, _old, new):
for _id, _ in list(self.multiselect.options):
if self.deaths.actual[_id].visible:
self.deaths.actual[_id].visible = False
self.deaths.predict[_id].visible = False
self.deaths.lower[_id].visible = False
self.deaths.upper[_id].visible = False
self.deaths.area[_id].visible = False
for _id in new:
if not self.deaths.actual[_id].visible:
_slice = self.deaths.data.loc[_id, :]
self.deaths.source[_id].data = ColumnDataSource.from_df(data=_slice)
self.deaths.actual[_id].visible = True
self.deaths.predict[_id].visible = True
self.deaths.lower[_id].visible = True
self.deaths.upper[_id].visible = True
self.deaths.area[_id].visible = True
def layout(self):
"""Build trend layout
Returns:
Bokeh Layout -- layout with cases, deaths and state selection
"""
_graphs = gridplot([self.cases.plot, self.deaths.plot], ncols=1,
plot_width=800 - self.multiselect.max_width,
plot_height=250, toolbar_location=None)
_layout = row(_graphs, self.multiselect)
return _layout