from bokeh.layouts import column, row
from bokeh.models import (
ColumnDataSource,
DataTable,
RangeTool,
ScientificFormatter,
StringFormatter,
TableColumn,
)
from bokeh.plotting import figure, show
from pastas.extensions import register_model_accessor
[docs]@register_model_accessor("bokeh")
class Bokeh:
"""Extension class for interactive bokeh figures for pastas Models.
Usage
-----
>>> ps.extensions.register_bokeh()
INFO: Registered bokeh plotting methods in Model class, e.g. `ml.bokeh.plot()`.
>>> fig = ml.bokeh.results()
>>> fig.write_html("results_figure.html")
Methods
-------
plot
plot oseries and model simulation, interactive version of `ml.plot()`
results
plot oseries, model simulation, contribution, step responses and parameters
table,interactive version of `ml.plots.results()`
Notes
-----
The `bokeh` extension is registered in the `Model` class by calling the
`register_bokeh()` function. To work in Juptyer notebooks, the
`bokeh.io.output_notebook()` function should be called before plotting. The `bokeh`
extension is not registered by default, and should be called explicitly. Check the
bokeh documentation for more information on how to interact with the plots.
"""
[docs] def __init__(self, model):
self._model = model
[docs] def plot(self, tmin=None, tmax=None, height=300, width=600, show_plot=True):
"""Plot the observations and model simulation.
Parameters
----------
tmin : pd.Timestamp, optional
start time for model simulation, by default None
tmax : pd.Timestamp, optional
end time for model simulation, by default None
height : int, optional
height of the plot, by default 500
width : int, optional
width of the plot, by default 800
show_plot : bool, optional
Show the plot (i.e., in Jupyter Notebooks), by default True
Returns
-------
p : bokeh.plotting.figure
Bokeh figure with the observations and model simulation.
Examples
--------
>>> ps.extensions.register_bokeh()
INFO: Registered bokeh plotting methods in Model class, e.g. `ml.bokeh.plot()`.
>>>
>>> fig = ml.bokeh.plot()
"""
data = self._model.get_output_series(tmin=tmin, tmax=tmax, split=False)
source = ColumnDataSource(data)
rsq = self._model.stats.rsq(tmin=tmin, tmax=tmax)
TOOLS = "zoom_in,zoom_out,reset,pan,xwheel_zoom,box_zoom,undo"
p = figure(
title="Pastas Model",
y_axis_label="Head",
x_axis_location=None,
tools=TOOLS,
width=width,
height=height,
x_axis_type="datetime",
toolbar_location="above",
)
p.scatter(
"index",
"Head_Calibration",
source=source,
legend_label="Observations",
color="black",
alpha=0.7,
)
p.line(
"index",
"Simulation",
source=source,
legend_label=r"Simulation (R2 = {:.2f})".format(rsq),
line_width=2,
)
p.legend.ncols = 2
if show_plot:
show(p)
return p
[docs] def results(self, tmin=None, tmax=None, height=500, width=800, show_plot=True):
"""Overview of the results of the pastas model.
Parameters
----------
tmin : pd.Timestamp, optional
start time for model simulation, by default None
tmax : pd.Timestamp, optional
end time for model simulation, by default None
height : int, optional
height of the plot, by default 500
width : int, optional
width of the plot, by default 800
show_plot : bool, optional
Show the plot (i.e., in Jupyter Notebooks), by default True
Returns
-------
grid : bokeh.layouts.column
Bokeh layout with the results of the pastas model.
Examples
--------
>>> ps.extensions.register_bokeh()
INFO: Registered bokeh plotting methods in Model class, e.g. `ml.bokeh.plot()`.
>>> fig = ml.bokeh.results()
"""
data = self._model.get_output_series(tmin=tmin, tmax=tmax, split=False)
ranges = data.max() - data.min()
ranges = ranges.drop([ranges.iloc[:2].idxmin(), "Noise"])
heights = (ranges / ranges.sum() * (height - 50)).values.astype(int)
source = ColumnDataSource(data)
rsq = self._model.stats.rsq(tmin=tmin, tmax=tmax)
TOOLS = "zoom_in,zoom_out,reset,pan,xwheel_zoom,box_zoom,undo"
p = figure(
title="Pastas Model",
y_axis_label="Head",
x_axis_location=None,
tools=TOOLS,
width=int(0.75 * width),
height=heights[0],
x_axis_type="datetime",
toolbar_location="above",
)
p.scatter(
"index",
"Head_Calibration",
source=source,
legend_label="Observations",
color="black",
alpha=0.7,
)
p.line(
"index",
"Simulation",
source=source,
legend_label=r"Simulation (R2 = {:.2f})".format(rsq),
)
p.legend.ncols = 2
# Residuals
res_plot = figure(
y_axis_label="Residuals",
toolbar_location=None,
tools=TOOLS,
x_range=p.x_range,
width=int(0.75 * width),
height=heights[2],
x_axis_type="datetime",
x_axis_location=None,
)
res_plot.scatter(
"index",
"Residuals",
source=source,
color="black",
alpha=0.7,
legend_label="Residuals",
)
res_plot.line(
"index",
"Residuals",
source=source,
color="black",
alpha=0.7,
legend_label="Residuals",
)
if self._model.settings["noise"]:
res_plot.line("index", "Noise", source=source, legend_label="Noise")
res_plot.scatter("index", "Noise", source=source, legend_label="Noise")
res_plot.legend.ncols = 2
# Parameter Table
df = ColumnDataSource(self._model.parameters.loc[:, ["optimal"]])
columns = [
TableColumn(
field="index",
title="Name",
formatter=StringFormatter(font_style="bold"),
),
TableColumn(
field="optimal",
title="Optimal",
formatter=ScientificFormatter(precision=2),
),
]
table = DataTable(
source=df,
columns=columns,
editable=False,
index_position=None,
width=int(0.25 * width),
height=heights[0] + heights[2] - 10,
)
left_column = [p, res_plot]
right_column = [table]
# Contributions
rfunc_plot = None
for i, smname in enumerate(self._model.stressmodels.keys(), start=2):
if i == int(len(self._model.stressmodels) + 1):
x_axis_location = "below"
else:
x_axis_location = None
contrib_plot = figure(
y_axis_label="Rise",
toolbar_location=None,
tools=TOOLS,
x_axis_location=x_axis_location,
width=int(0.75 * width),
height=heights[i],
x_axis_type="datetime",
x_range=p.x_range,
)
# xrange = False if rfunc_plot is None else rfunc_plot.x_range
rfunc_plot = figure(
x_axis_label=None,
toolbar_location=None,
x_axis_location=x_axis_location,
width=int(0.25 * width),
height=heights[i],
)
contrib_plot.line("index", smname, source=source)
response = self._model.get_step_response(smname)
rfunc_plot.line(response.index, response.values)
left_column.append(contrib_plot)
right_column.append(rfunc_plot)
select = figure(
height=50,
width=int(0.75 * width),
y_range=p.y_range,
x_axis_type="datetime",
y_axis_type=None,
tools="",
toolbar_location=None,
background_fill_color="#ffffff",
)
range_tool = RangeTool(x_range=p.x_range)
select.line("index", "Simulation", source=source)
select.add_tools(range_tool)
left_column.append(select)
layout = row(column(left_column), column(right_column))
grid = column(layout, width=width, height=height)
if show_plot:
show(grid)
return grid
