[
https://issues.apache.org/jira/browse/BEAM-7926?focusedWorklogId=335840&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-335840
]
ASF GitHub Bot logged work on BEAM-7926:
----------------------------------------
Author: ASF GitHub Bot
Created on: 30/Oct/19 00:37
Start Date: 30/Oct/19 00:37
Worklog Time Spent: 10m
Work Description: KevinGG commented on pull request #9741: [BEAM-7926]
Visualize PCollection
URL: https://github.com/apache/beam/pull/9741#discussion_r340388071
##########
File path:
sdks/python/apache_beam/runners/interactive/display/pcoll_visualization.py
##########
@@ -0,0 +1,269 @@
+#
+# Licensed to the Apache Software Foundation (ASF) under one or more
+# contributor license agreements. See the NOTICE file distributed with
+# this work for additional information regarding copyright ownership.
+# The ASF licenses this file to You under the Apache License, Version 2.0
+# (the "License"); you may not use this file except in compliance with
+# the License. You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+"""Module visualizes PCollection data.
+
+For internal use only; no backwards-compatibility guarantees.
+Only works with Python 3.5+.
+"""
+from __future__ import absolute_import
+
+import base64
+import logging
+from datetime import timedelta
+
+from pandas.io.json import json_normalize
+
+from apache_beam import pvalue
+from apache_beam.runners.interactive import interactive_environment as ie
+from apache_beam.runners.interactive import pipeline_instrument as instr
+
+try:
+ import jsons # pylint: disable=import-error
+ from IPython import get_ipython # pylint: disable=import-error
+ from IPython.core.display import HTML # pylint: disable=import-error
+ from IPython.core.display import Javascript # pylint: disable=import-error
+ from IPython.core.display import display # pylint: disable=import-error
+ from IPython.core.display import display_javascript # pylint:
disable=import-error
+ from IPython.core.display import update_display # pylint:
disable=import-error
+ from facets_overview.generic_feature_statistics_generator import
GenericFeatureStatisticsGenerator # pylint: disable=import-error
+ from timeloop import Timeloop # pylint: disable=import-error
+
+ if get_ipython():
+ _pcoll_visualization_ready = True
+ else:
+ _pcoll_visualization_ready = False
+except ImportError:
+ _pcoll_visualization_ready = False
+
+# 1-d types that need additional normalization to be compatible with DataFrame.
+_one_dimension_types = (int, float, str, bool, list, tuple)
+
+_DIVE_SCRIPT_TEMPLATE = """
+ document.querySelector("#{display_id}").data = {jsonstr};"""
+_DIVE_HTML_TEMPLATE = """
+ <script
src="https://cdnjs.cloudflare.com/ajax/libs/webcomponentsjs/1.3.3/webcomponents-lite.js";></script>
+ <link rel="import"
href="https://raw.githubusercontent.com/PAIR-code/facets/1.0.0/facets-dist/facets-jupyter.html";>
+ <facets-dive sprite-image-width="{sprite_size}"
sprite-image-height="{sprite_size}" id="{display_id}"
height="600"></facets-dive>
+ <script>
+ document.querySelector("#{display_id}").data = {jsonstr};
+ </script>"""
+_OVERVIEW_SCRIPT_TEMPLATE = """
+ document.querySelector("#{display_id}").protoInput =
"{protostr}";
+ """
+_OVERVIEW_HTML_TEMPLATE = """
+ <script
src="https://cdnjs.cloudflare.com/ajax/libs/webcomponentsjs/1.3.3/webcomponents-lite.js";></script>
+ <link rel="import"
href="https://raw.githubusercontent.com/PAIR-code/facets/1.0.0/facets-dist/facets-jupyter.html";>
+ <facets-overview id="{display_id}"></facets-overview>
+ <script>
+ document.querySelector("#{display_id}").protoInput =
"{protostr}";
+ </script>"""
+_DATAFRAME_PAGINATION_TEMPLATE = """
+ <script
src="https://ajax.googleapis.com/ajax/libs/jquery/2.2.2/jquery.min.js";></script>
+ <script
src="https://cdn.datatables.net/1.10.16/js/jquery.dataTables.js";></script>
+ <link rel="stylesheet"
href="https://cdn.datatables.net/1.10.16/css/jquery.dataTables.css";>
+ {dataframe_html}
+ <script>
+ $("#{table_id}").DataTable();
+ </script>"""
+
+
+def visualize(pcoll, dynamic_plotting_interval=None):
+ """Visualizes the data of a given PCollection. Optionally enables dynamic
+ plotting with interval in seconds if the PCollection is being produced by a
+ running pipeline or the pipeline is streaming indefinitely. The function
+ always returns immediately and is asynchronous when dynamic plotting is on.
+
+ If dynamic plotting enabled, the visualization is updated continuously until
+ the pipeline producing the PCollection is in an end state. The visualization
+ would be anchored to the notebook cell output area. The function
+ asynchronously returns a handle to the visualization job immediately. The
user
+ could manually do::
+
+ # In one notebook cell, enable dynamic plotting every 1 second:
+ handle = visualize(pcoll, dynamic_plotting_interval=1)
+ # Visualization anchored to the cell's output area.
+ # In a different cell:
+ handle.stop()
+ # Will stop the dynamic plotting of the above visualization manually.
+ # Otherwise, dynamic plotting ends when pipeline is not running anymore.
+
+ If dynamic_plotting is not enabled (by default), None is returned.
+ """
+ if not _pcoll_visualization_ready:
+ return None
+ pv = PCollectionVisualization(pcoll)
+ pv.display_facets()
+
+ if dynamic_plotting_interval:
+ # Disables the verbose logging from timeloop.
+ logging.getLogger('timeloop').disabled = True
+ tl = Timeloop()
+
+ def dynamic_plotting(pcoll, pv, tl):
+ @tl.job(interval=timedelta(seconds=dynamic_plotting_interval))
+ def continuous_update_display(): # pylint: disable=unused-variable
+ # Always creates a new PCollVisualization instance when the
+ # PCollection materialization is being updated and dynamic
+ # plotting is in-process.
+ updated_pv = PCollectionVisualization(pcoll)
+ updated_pv.display_facets(updating_pv=pv)
+ if ie.current_env().is_terminated(pcoll.pipeline):
+ try:
+ tl.stop()
+ except RuntimeError:
+ # The job can only be stopped once. Ignore excessive stops.
+ pass
+
+ tl.start()
+ return tl
+
+ return dynamic_plotting(pcoll, pv, tl)
+ return None
+
+
+class PCollectionVisualization(object):
+ """A visualization of a PCollection.
+
+ The class relies on creating a PipelineInstrument w/o actual instrument to
+ access current interactive environment for materialized PCollection data at
+ the moment of self instantiation through cache.
+ """
+
+ def __init__(self, pcoll):
+ assert _pcoll_visualization_ready, (
+ 'Dependencies for PCollection visualization are not available. Please '
+ 'use `pip install apache-beam[interactive]` to install necessary '
+ 'dependencies and make sure that you are executing code in an '
+ 'interactive environment such as a Jupyter notebook.')
+ assert isinstance(pcoll, pvalue.PCollection), (
+ 'pcoll should be apache_beam.pvalue.PCollection')
+ self._pcoll = pcoll
+ # This allows us to access cache key and other meta data about the pipeline
+ # whether it's the pipeline defined in user code or a copy of that
pipeline.
+ # Thus, this module doesn't need any other user input but the PCollection
+ # variable to be visualized. It then automatically figures out the pipeline
+ # definition, materialized data and the pipeline result for the execution
+ # even if the user never assigned or waited the result explicitly.
+ # With only the constructor of PipelineInstrument, any interactivity
related
+ # pre-process or instrument is not triggered for performance concerns.
+ self._pin = instr.PipelineInstrument(pcoll.pipeline)
+ self._cache_key = self._pin.cache_key(self._pcoll)
+ self._dive_display_id = 'facets_dive_{}_{}'.format(self._cache_key,
+ id(self))
+ self._overview_display_id = 'facets_overview_{}_{}'.format(self._cache_key,
+ id(self))
+ self._df_display_id = 'df_{}_{}'.format(self._cache_key, id(self))
+
+ def display_facets(self, updating_pv=None):
+ """Displays the visualization through IPython.
+
+ Args:
+ updating_pv: A PCollectionVisualization object. When provided, the
+ display_id of each visualization part will inherit from the initial
+ display of updating_pv and only update that visualization web element
+ instead of creating new ones.
+
+ The visualization has 3 parts: facets-dive, facets-overview and paginated
+ data table. Each part is assigned an auto-generated unique display id
+ (the uniqueness is guaranteed throughout the lifespan of the PCollection
+ variable).
+ """
+ # Ensures that dive, overview and table render the same data because the
+ # materialized PCollection data might being updated continuously.
+ data = self._to_dataframe()
+ if updating_pv:
+ self._display_dive(data, updating_pv._dive_display_id)
+ self._display_overview(data, updating_pv._overview_display_id)
+ self._display_dataframe(data, updating_pv._df_display_id)
+ else:
+ self._display_dive(data)
+ self._display_overview(data)
+ self._display_dataframe(data)
+
+ def _display_dive(self, data, update=None):
+ sprite_size = 32 if len(data.index) > 50000 else 64
+ jsonstr = data.to_json(orient='records')
+ if update:
+ script = _DIVE_SCRIPT_TEMPLATE.format(display_id=update,
+ jsonstr=jsonstr)
+ display_javascript(Javascript(script))
+ else:
+ html = _DIVE_HTML_TEMPLATE.format(display_id=self._dive_display_id,
+ jsonstr=jsonstr,
+ sprite_size=sprite_size)
+ display(HTML(html))
Review comment:
For ipython terminal, the users would see something similar to below:
```
<IPython.core.display.HTML object>
<IPython.core.display.HTML object>
<IPython.core.display.HTML object>
```
I think the `show`(`visualization`) here heavily depends on HTML/JS/CSS to
make sense (because it tries to load the whole PCollection cache into memory
for rendering), so if the kernel is connected to a frontend that supports
these, they should all support the visualization.
We can add a default pretty print that works similar to existing
`print(result.get_list(pcoll))` and it's nothing close to what current
`visualization()` does. However, the kernel doesn't know whether it is
connected to a frontend / what kinds of frontends, so it's either always
pretty-print or not doing it at all.
I would propose limit the work here for the scenarios when the kernel is
actually connected to a frontend. And data visualization in a plain ipython
terminal is not what we want to achieve here since one cannot interact with the
rendering in a terminal easily (with mouse and etc.).
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Issue Time Tracking
-------------------
Worklog Id: (was: 335840)
Time Spent: 18h 10m (was: 18h)
> Visualize PCollection with Interactive Beam
> -------------------------------------------
>
> Key: BEAM-7926
> URL: https://issues.apache.org/jira/browse/BEAM-7926
> Project: Beam
> Issue Type: New Feature
> Components: runner-py-interactive
> Reporter: Ning Kang
> Assignee: Ning Kang
> Priority: Major
> Time Spent: 18h 10m
> Remaining Estimate: 0h
>
> Support auto plotting / charting of materialized data of a given PCollection
> with Interactive Beam.
> Say an Interactive Beam pipeline defined as
> p = create_pipeline()
> pcoll = p | 'Transform' >> transform()
> The use can call a single function and get auto-magical charting of the data
> as materialized pcoll.
> e.g., visualize(pcoll)
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
