Optview dash (#200)

* Updated requirements to support dash

* Created new dash optview version

* Revert "Updated requirements to support dash"

This reverts commit 64e5a3d.

* Revert "Created new dash optview version"

This reverts commit e6466e3.

* Added env to gitignore

* Added OptView folder

* Documented  addMaxMinTraces

* Fixed linting errors

* Enabled space for longer var names

* undo commented out NSGA2

* deleted old dash file

* moved OptViewDash files

* added optview dash to setup.py

* fixed some typos

* moved OptViewIcon to assets directory

* updated CRLF to LF endings

* added tests for optview and optview_dash

* updated travis

* removed optview test since cannot test in headless

* forgot to add reason for skipTest

* added docs

* Revert "updated CRLF to LF endings"

This reverts commit 19f0346.

* Delete .flake8

* Revert "Delete .flake8"

This reverts commit 89813cc.

Co-authored-by: Marco Mangano <36549388+marcomangano@users.noreply.github.com>
Co-authored-by: Neil Wu <neilwu@umich.edu>
Co-authored-by: Neil Wu <602725+nwu63@users.noreply.github.com>

.gitignore

@@ -7,4 +7,5 @@ build/*
 *.hst
 *.gz
 *.c
-*.cpp
+*.cpp
+env

.travis.yml

@@ -23,7 +23,7 @@ install:
       docker exec -it app /bin/bash -c "cp -r \$HOME/NLPQLP/* $NLPQLP_DIR/ && cp -r \$HOME/SNOPT/* $SNOPT_DIR/";
     fi
   # Build and install
-  - docker exec -it app /bin/bash -c ". $BASHRC && cd $DOCKER_WORKING_DIR && pip install -e ."
+  - docker exec -it app /bin/bash -c ". $BASHRC && cd $DOCKER_WORKING_DIR && pip install -e .[optview]"
 
 script:
   - |

doc/install.rst

@@ -18,6 +18,7 @@ Simply comment out the corresponding lines in ``pyoptsparse/pyoptsparse/setup.py
 The corresponding lines in ``pyoptsparse/__init__.py`` must be commented out as well.
 
 Python dependencies are automatically handled by ``pip``, so they do not need to be installed separately.
+The only exception is ``numpy``, which is required as part of the build process and therefore must be present before installing.
 
 .. note::
   * In Linux, the python header files (python-dev) are also required.
@@ -46,6 +47,16 @@ It is also possible to install pyOptSparse by calling ``python setup.py install`
 
 For those who intend to use pyOptSparse with IPOPT, OpenMDAO developers provide a `bash script <https://github.com/OpenMDAO/build_pyoptsparse>`_ that simplifies the installation of the optimizer with different linear solvers.
 
+.. _install_optview:
+
+Installing OptView
+------------------
+OptView and OptView-Dash have separate dependencies that must be installed.
+To install pyOptSparse including those dependencies, run::
+
+    pip install .[optview]
+
+
 Testing
 -------
 pyOptSparse provides a set of unit and regression tests to verify the installation.

doc/postprocessing.rst

@@ -1,8 +1,9 @@
 Post-processing
 ===============
-There are two post-processing utilities that are provided with pyOptSparse.
+There are three post-processing utilities that are provided with pyOptSparse.
 
 - OptView is a GUI designed to quickly and interactively visualize optimization histories
+- OptView-Dash is a `Dash <https://plotly.com/dash/>`_ implementation of OptView
 - ``History`` is a Python class that can be used to read in the history file, and provide API for programmatically extracting data.
 
 OptView
@@ -23,11 +24,8 @@ Requirements
       \-axisartist (OptView)
     numpy (OptView)
 
-If you are successfully running pyOptSparse, these packages are most likely
-already installed.
-
-Although not necessary for most usage, the ``dill`` package is needed
-if you wish to save an editable version of the graph produced in ``OptView``.
+For installation instructions, see :ref:`install_optview`.  
+Although not necessary for most usage, the ``dill`` package is needed if you wish to save an editable version of the graph produced in ``OptView``.
 ``dill`` can be installed via ``pip`` in a terminal using::
 
     pip install dill
@@ -40,18 +38,15 @@ Usage
 
     optview histFile
 
-Here, ``histFile`` is the name of the history file to be examined
-(default is ``opt_hist.hst``).
+Here, ``histFile`` is the name of the history file to be examined (default is ``opt_hist.hst``).
 
 
-Additionally, you can open multiple history files in the same ``OptView`` instance
-by calling them via the command line::
+Additionally, you can open multiple history files in the same ``OptView`` instance by calling them via the command line::
 
     optview histFile1 histFile2 histFile3
 
-Each file's contents will be loaded into ``OptView`` with a flag appended to the end
-of each variable or function name corresponding to the history file. The first one
-listed will have '_A' added to the name, the second will have '_B' added, etc.
+Each file's contents will be loaded into ``OptView`` with a flag appended to the end of each variable or function name corresponding to the history file.
+The first one listed will have '_A' added to the name, the second will have '_B' added, etc.
 There is currently no limit to the number of history files than can be loaded.
 
 Optionally, if you want to save the generated figures, there is an optional argument::
@@ -72,23 +67,18 @@ Features
     * saving the figure to an image or pickling it for later formatting
     * refreshing the optimization history on the fly
 
-Although some of these are self-explanatory, the layout and usage of ``OptView``
-will be explained below.
+Although some of these are self-explanatory, the layout and usage of ``OptView`` will be explained below.
 
 GUI Layout
 ~~~~~~~~~~
 The window is divided into two sections.
-The top is the canvas where the figure and graphs will be produced,
-while the bottom grayed section contains user-selectable options.
+The top is the canvas where the figure and graphs will be produced, while the bottom grayed section contains user-selectable options.
 Here, we will focus on the user options.
 
-The selectable variables are contained on the left hand
-side of the options panel in scrollable listboxes.
-You can select multiple items from the listboxes using the normal selection
-operators such as control and shift.
-If a selected variable is an array, a third listbox should appear on the
-right hand side of the options panel, allowing you to select specific
-sub-variables within the single array variable.
+The selectable variables are contained on the left hand side of the options panel in scrollable listboxes.
+You can select multiple items from the listboxes using the normal selection operators such as control and shift.
+If a selected variable is an array, a third listbox should appear on the right hand side of the options panel,
+allowing you to select specific sub-variables within the single array variable.
 
 There are three main options when selecting how to produce the graph(s):
 
@@ -97,8 +87,7 @@ There are three main options when selecting how to produce the graph(s):
     * Stacked plots - each variable gets its own individual plot and the set is stacked vertically
 
 Most checkbox options should play well with any of these three main options,
-though there are known issues with using the 'multiple axes'
-option and delta values or for displaying arrays.
+though there are known issues with using the 'multiple axes' option and delta values or for displaying arrays.
 
 There are seven checkbox options:
 
@@ -124,10 +113,23 @@ Lastly, there are some miscellaneous features:
     * Hoverable tooltips when the cursor is on a plot line
     * A variable called `actual_iteration_number` that gives a translation between history file iteration number and run file iteration number. This is especially useful for debugging specific steps of an optimization or comparing values across different histories.
 
-More features are being developed on an as-needed basis.
-Feel free to edit the code as you see fit and submit a pull request if you
-would like to see a feature added.
-Alternatively, you can submit an issue ticket to discuss possible features.
+
+OptView-Dash
+------------
+This is a Dash_ implementation of OptView, and has many of the same features offered by OptView.
+For installation instructions, see :ref:`install_optview`.
+To run, use this command::
+
+    optview_dash <filename>
+
+Similar to OptView, you can invoke it with multiple history files.
+To view the dash app, you will have to manually open the server in your browser that is listed in the terminal after running the above command.
+
+Auto-refresh: This follows the same functionality as OptView, allowing you to see the changes of an optimization as it is running.
+
+-  If you toggle this checklist button, it will cause the program to default update every 10 seconds, however you may modify this refresh rate using the input box underneath
+-  Make sure to toggle off this button when you are done or the optimization is complete so it does not add lag.
+-  This feature also works with multiple history files/optimizations running!
 
 Directly Accessing the History Object
 -------------------------------------

pyoptsparse/postprocessing/OptView.py

@@ -14,18 +14,9 @@
 import os
 import argparse
 import shelve
-
 import sys
-
-major_python_version = sys.version_info[0]
-
-if major_python_version == 2:
-    import tkFont
-    import Tkinter as Tk
-else:
-    import tkinter as Tk
-    from tkinter import font as tkFont
-
+import tkinter as Tk
+from tkinter import font as tkFont
 import re
 import warnings
 
@@ -47,7 +38,6 @@
     pass
 import numpy as np
 from sqlitedict import SqliteDict
-import traceback
 from .OptView_baseclass import OVBaseClass
 
 
@@ -69,7 +59,7 @@ def __init__(self, histList, outputDir, figsize):
         try:
             icon_dir = os.path.dirname(os.path.abspath(__file__))
             icon_name = "OptViewIcon.gif"
-            icon_dir_full = os.path.join(icon_dir, icon_name)
+            icon_dir_full = os.path.join(icon_dir, "assets", icon_name)
             img = Tk.PhotoImage(file=icon_dir_full)
             self.root.tk.call("wm", "iconphoto", self.root._w, img)
         except:  # bare except because error is not in standard Python