The validation module provides the capability to perform integrity checks on an entire collection of Bokeh models.
To create a Bokeh visualization, the central task is to assemble a collection model objects from bokeh.plotting into a graph that represents the scene that should be created in the client. It is possible to to this “by hand”, using the model objects directly. However, to make this process easier, Bokeh provides higher level interfaces such as bokeh.plotting for users.
These interfaces automate common “assembly” steps, to ensure a Bokeh object graph is created in a consistent, predictable way. However, regardless of what interface is used, it is possible to put Bokeh models together in ways that are incomplete, or that do not make sense in some way.
To assist with diagnosing potential problems, Bokeh performs a validation step when outputting a visualization for display. This module contains error and warning codes as well as helper functions for defining validation checks.
One use case for warnings is to loudly point users in the right direction when they accidentally do something that they probably didn’t mean to do - this is the case for EMPTY_LAYOUT for instance. Since warnings don’t necessarily indicate misuse, they are configurable. To silence a warning, use the silence function provided.
>>> from bokeh.core.validation import silence >>> from bokeh.core.validation.warnings import EMPTY_LAYOUT >>> silence(EMPTY_LAYOUT, True)
These define the standard error codes and messages for Bokeh validation checks.
A glyph has a property set to a field name that does not correspond to any column in the GlyphRenderer’s data source.
GlyphRenderer
A GlyphRenderer has no glyph configured.
A GlyphRenderer has no data source configured.
A Plot is missing one or more required default ranges (will result in blank plot).
Plot
Google Maps API now requires an API key for all use. See https://developers.google.com/maps/documentation/javascript/get-api-key for more information on how to obtain your own, to use for the api_key property of your Google Map plot .
api_key
All data_sources on LegendItem.renderers must match when LegendItem.label is type field.
LegendItem.renderers
MercatorTicker and MercatorTickFormatter``models must have their ``dimension property set to 'lat' or 'lon'.
MercatorTicker
MercatorTickFormatter``models must have their ``dimension
'lat'
'lon'
A Scale on is missing one or more required default scales (will result in blank plot).
Scale
A Scale type is incompatible with one or more ranges on the same plot dimension (will result in blank plot).
A GlyphRenderer has a CDSView whose source doesn’t match the GlyphRenderer’s data source.
CDSView
The GraphSource is incorrectly configured.
GraphSource
Map plots can only support Range1d types, not data ranges.
Range1d
The PointDrawTool renderers may only reference XYGlyph models.
PointDrawTool
XYGlyph
The BoxEditTool renderers may only reference Rect glyph models.
BoxEditTool
Rect
The PolyDrawTool renderers may only reference MultiLine and Patches glyph models.
PolyDrawTool
MultiLine
Patches
The PolyEditTool renderers may only reference MultiLine and Patches glyph models.
PolyEditTool
The PolyEditTool vertex_renderer may only reference XYGlyph models.
The RangeTool must have at least one of x_range or y_range configured
RangeTool
x_range
y_range
FactorRange must specify a unique list of categorical factors for an axis.
FactorRange
An extra range name is configured with a name that does not correspond to any range.
noUiSlider most have a nonequal start and end.
noUiSlider
Expected min_width <= width <= max_width
Expected min_height <= height <= max_height
CDSView filters are not compatible with glyphs with connected topology such as Line or Patch.
The LineEditTool renderers may only reference MultiLine and Line glyph models.
LineEditTool
Line
The LineEditTool intersection_enderer may only reference LineGlyph models.
LineGlyph
The same model can’t be used multiple times in a layout.
Indicates that a custom error check has failed.
These define the standard warning codes and messages for Bokeh validation checks.
A Plot object has no renderers configured (will result in a blank plot).
A layout model has no children (will result in a blank layout).
Each component can be rendered in only one place, can’t be both a root and in a layout.
Indicates that a custom warning check has failed.
These helper functions can be used to perform integrity checks on collections of Bokeh models, or mark methods on Models as warning or error checks.
check_integrity
Apply validation and integrity checks to a collection of Bokeh models.
models (seq[Model]) – a collection of Models to test
None
This function will emit log warning and error messages for all error or warning conditions that are detected. For example, layouts without any children will trigger a warning:
>>> empty_row = Row >>> check_integrity([empty_row]) W-1002 (EMPTY_LAYOUT): Layout has no children: Row(id='2404a029-c69b-4e30-9b7d-4b7b6cdaad5b', ...)
silence
Silence a particular warning on all Bokeh models.
warning (Warning) – Bokeh warning to silence
silence (bool) – Whether or not to silence the warning
A set containing the all silenced warnings
This function adds or removes warnings from a set of silencers which is referred to when running check_integrity. If a warning is added to the silencers - then it will never be raised.
>>> from bokeh.core.validation.warnings import EMPTY_LAYOUT >>> bokeh.core.validation.silence(EMPTY_LAYOUT, True) {1002}
To turn a warning back on use the same method but with the silence argument set to false
>>> bokeh.core.validation.silence(EMPTY_LAYOUT, False) set()
error
Decorator to mark a validator method for a Bokeh error condition
code_or_name (int or str) – a code from bokeh.validation.errors or a string label for a custom check
bokeh.validation.errors
decorator for Bokeh model methods
callable
The function that is decorated should have a name that starts with _check, and return a string message in case a bad condition is detected, and None if no bad condition is detected.
_check
Examples:
The first example uses a numeric code for a standard error provided in bokeh.validation.errors. This usage is primarily of interest to Bokeh core developers.
from bokeh.validation.errors import REQUIRED_RANGES @error(REQUIRED_RANGES) def _check_no_glyph_renderers(self): if bad_condition: return "message"
The second example shows how a custom warning check can be implemented by passing an arbitrary string label to the decorator. This usage is primarily of interest to anyone extending Bokeh with their own custom models.
@error("MY_CUSTOM_WARNING") def _check_my_custom_warning(self): if bad_condition: return "message"
warning
The first example uses a numeric code for a standard warning provided in bokeh.validation.warnings. This usage is primarily of interest to Bokeh core developers.
bokeh.validation.warnings
from bokeh.validation.warnings import MISSING_RENDERERS @warning(MISSING_RENDERERS) def _check_no_glyph_renderers(self): if bad_condition: return "message"
@warning("MY_CUSTOM_WARNING") def _check_my_custom_warning(self): if bad_condition: return "message"