orchid-python-api 2021.1.399

Defines and implements the Python API for Orchid*. (*Orchid is a mark of Reveal Energy Services, Inc.)

Introduction

This project defines the implementation of the Python API for Orchid*.

(*Orchid in a mark of Revel Energy Services. Inc.)

Specifically, the orchid package exposes the Orchid API to Python applications and to the Python REPL.

Getting Started

Create a virtual environment

Examples

Additionally, this project installs four examples in the examples directory of the orchid-python-api package:

• plot_trajectories.ipynb
• plot_monitor_curves.ipynb
• plot_treatment.ipynb
• completion_analysis.ipynb

The first three notebooks plot:

• The well trajectories for a project
• The monitor curves for a project
• The treatment curves (pressure, slurry rate and concentration) for a specific stage of a well in a project

Additionally, the notebook, completion_analysis.ipynb, provides a more detailed analysis of the completion performed on two different wells in a project.

To use these examples:

• You may need to configure the Orchid Python API to find the Orchid installation

• You must configure the Orchid Python API to find the Orchid training data

• You may need to view the Orchid API configuration details

• You may want to invoke the command, copy_orchid_examples

  This command copies the example files into an optionally specified (virtual environment) directory. (The default destination is your current working directory.) Note that this command is a command-line script that runs in a console or terminal. Additionally, this command supports a help flag (-h / --help) to provide you with help on running this command.

End-user preparation

We recommend the use of virtual environments to use the Orchid Python API. This choice avoids putting Orchid-specific packages in your system Python environment.

You have several options to create and manage virtual environments: venv, pipenv, poetry, and conda. The venv is available as a standard Python package and is a spartan tool to manage environments. poetry is a tool targeting developers but can be used by end-users. Our recommended tool is pipenv. It provides a good balance between venv and poetry. Remember, both pipenv and poetry must be installed in your Python environment separately from Python itself, but can be installed using pip. Finally, conda

is an open source package management system and environment management system that runs on Windows, macOS and Linux. Conda quickly installs, runs and updates packages and their dependencies. Conda easily creates, saves, loads and switches between environments on your local computer. It was created for Python programs, but it can package and distribute software for any language.

Although we currently recommend pipenv, because we believe many of our users use conda (either Anaconda or Miniconda), we have a section for installing the orchid-python-api in a conda virtual environment.

Using any of pipenv, venv or poetry, your first step is to create a directory for your project. Then, change into your project directory.

We recommend the use of pipenv. This environment hides a number of details involved in managing a virtualenv and yet provides a fairly simple interface. We will assume in this document that you are using pipenv.

Step-by-step install

• Install python 3.8 by following these instructions. To ensure access from the command line, be sure to select the "Add Python 3.x to PATH" option on the installer start page.
• Open a console using either powershell or the Windows console.
• Create a directory for the virtual environment. We will symbolically call it /path/to/orchid-virtualenv.
• Change the current working directory by executing, chdir /path/to/orchid-virtualenv.
• Create an empty virtual environment by running pipenv install.
• Activate the virtual environment by running pipenv shell
• Install Orchid by running pip install orchid-python-api.

Step-by-step conda install

• Install Anaconda or Miniconda following the corresponding instructions for your operating system.
• If installing on Windows, the installer will present this screen. We have seen no need to install Anaconda / Miniconda on your PATH. Although we do not disagree with option to register the Anaconda / Miniconda version of Python as your default Python executable, in some situations, accepting this choice can cause problems.
• Since we will be using both conda install and pip install to install packages, read the article, Using Pip in a Conda Environment. Our subsequent instructions assume you have read this article and have chosen how you wish to manage these two package installers together.

The following instructions assume that you will use the simple (put perhaps not scalable) process of creating the conda virtual environment with all packages you want to use available in the Anaconda/Miniconda ecosystem and, within that virtual environment, use pip to install orchid-python-api.

• Open an Anaconda Powershell console.
• Optionally create a directory for your work.
  • We symbolically call it /path/to/orchid-virtualenv.
  • Change to the current working directory by executing chdir /path/to/orchid-virtualenv.
• Create an empty virtual environment by running conda create --name <your-virtualenv-name> python=3.8.
• Activate the virtual environment by running conda activate <your-virtualenv_name>
• Install Orchid by running pip install orchid-python-api.

Configure the Orchid Python API

The Orchid Python API requires a licensed Orchid installation on your workstation. Depending on the details of the installation, you may need to configure the Orchid Python API to refer to different locations.

Using the fallback configuration

If you installed the latest version Orchid using the installation defaults, and you installed the orchid-python-api , you need to take no additional steps to configure the Orchid Python API to find this installation. For your information, the default installation location is, %ProgramFiles%\Reveal Energy Services, Inc\Orchid. The Orchid Python API uses its version to find and use the corresponding version of Orchid.

Using an environment variable

This mechanism is perhaps the easiest procedure to create an Orchid Python API configuration that changes rarely and is available to all your tools. It works best with a system restart. (Environment variables can be made available for a narrow set of tools on your system or available to all your tools depending on arcane technical rules that you need not understand.)

To use environment variables to configure the Orchid Python API, you will need to create the environment variable ORCHID_ROOT and set its value to the root Orchid installation directory. (For your information, the version-specific Orchid binary files, .exe's and .dll's should be in a subdirectory of ORCHID_ROOT with a name like Orchid-2020.4.232.)

This document assumes you want to create a long-term configuration that survives a system restart and is available to all your tools. Symbolically, this document will refer to the root of the Orchid installation as /path/to/orchid-installation.

To create the required environment variable, enter the search term "environment variables" in the Windows-10 search box and select the item named, "Edit environment variables for your account." The system will then present your with the "Environment Variables" dialog. Under the section named "User variables for <your.username>", click the "New" button. In the "Variable name" text box, enter "ORCHID_ROOT". (These two words are separated by the underscore symbol.)

Navigate to the "Variable Value" text box. Click the "Browse Directory" button to select the directory into which Orchid is installed, /path/to/orchid-installation. This action pastes the directory name into the "Variable Value" text box. Verify that the directory is correct, and then click "OK". Verify that you see the name ORCHID_ROOT with the correct value in the "User variables for <your.username>" list. Finally, click "OK" to dismiss the "Environment Variables" dialog.

Although you have now created the ORCHID_ROOT environment variable with the appropriate value, only "new" tools can now use that variable. However, the details of "new" is technical and may not correspond to your what you expect. If you understand these details, you can jump to Verify Installation. If you are not confident of these details, restart your system before proceeding to Verify Installation.

Using a configuration file

Another option to configure the Orchid Python API is by creating a configuration file. A configuration file is easier to change than an environment variable and does not require a system restart to work best. However, it requires more knowledge and work on your part. In general, a configuration file is better if your requirements change "often". For example, if you are working with multiple, side-by-side Orchid versions and Orchid Python API versions, you may find it faster and easier to create a configuration file once and change it as you change Orchid / Orchid Python API versions.

To create a configuration file used by the Orchid Python API, you create a file named python_api.yaml and put it in the directory, /path/to/home-directory/.orchid, where /path/to/home-directory is a symbolic reference to your home directory. Technically, the format of the file is YAML ("YAML Ain't Markup Language"), a "human friendly data serialization standard". (For technical details, visit the website. For a gentler introduction, visit the Wikipedia entry or read / watch on of the many YAML introductions / tutorials.)

Because these articles describe YAML generally, they do not describe the details of the YAML document expected by the Orchid Python API. We, however, distribute an example file name python_api.yaml.example in each installed orchid-python-api package. Assuming you created a virtual environment as described in Step-by step install, you can find this example file, python_api.yaml.example, in the directory, /path/to/orchid-virtualenv/Lib/site-packages/orchid_python_api/examples.

To use this configuration file as an example:

• Copy the file to the expected location. For example, assuming the symbolic names referenced above, execute copy /path/to/orchid-virtualenv/Lib/site-packages/orchid_python_api/examples/python_api.yaml.example /path/to/home-directory/.orchid/python_api.yaml
• Edit the copied file, /path/to/home-directory/.orchid/python_api.yaml, using your favorite text editor.

The example file, contains comments, introduced by a leading octothorpe character (#, number sign, or hash), that describe the information expected by the Orchid Python API. In summary, you'll need to provide a value for the 'orchid' > 'root' key that contains the pathname of the directory containing the Orchid binaries corresponding to the installed version of the orchid-python-api package.

If you want to ensure your configuration is correct, view the Orchid API configuration details.

Configure the Orchid training data

The Orchid Python API requires a licensed Orchid installation on your workstation. However, configuring the Orchid Python API to find the Orchid training data is only needed to run the example Jupyter notebooks.

Using an environment variable

This mechanism is perhaps the easiest procedure to create an Orchid Python API configuration that changes rarely and is available to all your tools. It works best with a system restart. (Environment variables can be made available for a narrow set of tools on your system or available to all your tools depending on arcane technical rules that you need not understand.)

To use environment variables to configure the Orchid Python API to find the Orchid training data, you will need to create the environment variable ORCHID_TRAINING_DATA and set its value to the location of the Orchid training data.

This document assumes you want to create a long-term configuration that survives a system restart and is available to all your tools. Symbolically, this document will refer to the Orchid training data location as /path-to/orchid/training-data.

To create the required environment variable, enter the search term "environment variables" in the Windows-10 search box and select the item named, "Edit environment variables for your account." The system will then present your with the "Environment Variables" dialog. Under the section named "User variables for <your.username>", click the "New" button. In the "Variable name" text box, enter "ORCHID_TRAINING_DATA". (These two words are separated by the underscore symbol.)

Navigate to the "Variable Value" text box. Click the "Browse Directory" button to select the directory containing the Orchid training data, /path-to/orchid/training-data. This action pastes the directory name into the "Variable Value" text box. Verify that the directory is correct, and then click "OK". Verify that you see the name ORCHID_TRAINING_DATA with the correct value in the "User variables for <your.username>" list. Finally, click "OK" to dismiss the "Environment Variables" dialog.

Although you have now created the ORCHID_ROOT environment variable with the appropriate value, only "new" tools can now use that variable. However, the details of "new" is technical and may not correspond to your what you expect. If you understand these details, you can jump to Verify Installation. If you are not confident of these details, restart your system before proceeding to Verify Installation.

Using a configuration file

Another option to configure the Orchid Python API to find the Orchid training data is by creating a configuration file. A configuration file is easier to change than an environment variable and does not require a system restart to work best. However, it requires more knowledge and work on your part. In general, a configuration file is better if your requirements change "often". For example, if you are working with multiple, side-by-side Orchid versions and Orchid Python API versions, you may find it faster and easier to create a configuration file once and change it as you change Orchid / Orchid Python API versions.

To create a configuration file used by the Orchid Python API, you create a file named python_api.yaml and put it in the directory, /path/to/home-directory/.orchid, where /path/to/home-directory is a symbolic reference to your home directory. Technically, the format of the file is YAML ("YAML Ain't Markup Language"), a "human friendly data serialization standard". (For technical details, visit the website. For a gentler introduction, visit the Wikipedia entry or read / watch on of the many YAML introductions / tutorials.)

Because these articles describe YAML generally, they do not describe the details of the YAML document expected by the Orchid Python API. We, however, distribute an example file name python_api.yaml.example in each installed orchid-python-api package. Assuming you created a virtual environment as described in Step-by step install, you can find this example file, python_api.yaml.example, in the directory, /path/to/orchid-virtualenv/Lib/site-packages/orchid_python_api/examples.

To use this configuration file as an example:

• Copy the file to the expected location. For example, assuming the symbolic names referenced above, execute copy /path/to/orchid-virtualenv/Lib/site-packages/orchid_python_api/examples/python_api.yaml.example /path/to/home-directory/.orchid/python_api.yaml
• Edit the copied file, /path/to/home-directory/.orchid/python_api.yaml, using your favorite text editor.

The example file, contains comments, introduced by a leading octothorpe character (#, number sign, or hash), that describe the information expected by the Orchid Python API. In summary, you'll need to provide a value for the 'orchid' > 'training_data' key that contains the pathname of the directory containing the Orchid binaries corresponding to the installed version of the orchid-python-api package.

If you want to ensure your configuration is correct, view the Orchid API configuration details.

Verify installation

Jupyter lab

• In your activated virtual environment, run jupyter lab to open a browser tab.
• In the first cell, enter import orchid.
• Run the cell.
• Wait patiently.

The import should complete with no errors.

Python REPL

• In your activated virtual environment, run python to open a REPL.
• Enter import orchid.
• Wait patiently.

The import should complete with no errors.

Run orchid examples

• Navigate to the directory associated with the virtual environment
• Run python </path/to/virtualenv/Lib/site-packages/copy_orchid_examples.py
• If the script reports that it skipped notebooks, repeat the command with an additional argument:
  python </path/to/virtualenv/Lib/site-packages/copy_orchid_examples.py --overwrite
• Verify that the current directory has four notebooks:
  • plot_trajectories.ipynb
  • plot_monitor_curves.ipynb
  • plot_treatment.ipynb
  • completion_analysis.ipynb
• The notebooks, as installed, "symbolically" reference the Orchid training data. To resolve this "symbolic reference", configure the Orchid Python API to find the Orchid training data.
• Activate your virtual environment by pipenv shell if not already activated
• Open Jupyter by running jupyter lab in the shell
• Within Jupyter, Run the notebook, plot_trajectories.ipynb 1. Open notebook 2. Run all cells of notebook 3. Wait patiently 4. Verify that no exceptions occurred
  • Repeat for remaining notebooks:
    • plot_monitor_curves.ipynb
    • plot_treatment.ipynb
    • completion_analysis.ipynb

View Orchid Configuration Details

To "debug" the Orchid Python API configuration, perform the following steps:

• Change to the directory associated with your Python virtual environment.
• If necessary, activate the virtual environment.
• Within that virtual environment, invoke Python. It is important to create a new REPL so that you start with a "clean" environment.
• Within the Python REPL, execute the following commands.

  import logging
  logging.basicConfi(level=logging.DEBUG)
  import orchid
  

Enabling logging before importing is critical. If you have already imported orchid, the simplest solution is to close this REPL and create another, "clean" REPL.

You should see output like the following:

DEBUG:orchid.configuration:fallback configuration={'orchid': {'root': 'C:\\Program Files\\Reveal Energy Services, Inc\\Orchid\\Orchid-2020.4.361'}}
DEBUG:orchid.configuration:file configuration={'orchid': {'root': 'c:\\path-to\\bin\\x64\\Debug\\net48', 'training_data ': 'c:\\path-to\\installed-training-data'}}
DEBUG:orchid.configuration:environment configuration = {'orchid': {'root': 'c:\\another\\path-to\bin\\x64\\Debug\\net48'}}
DEBUG:orchid.configuration:result configuration={'orchid': {'root': 'c:\\another\\path-to\bin\\x64\\Debug\\net48'}}

This output describes four details of the configuration.

Configuration	Explanation
result	The configuration used by the Orchid Python API
fallback	The always available configuration
file	The configuration specified in your configuration file
environment	The configuration specified using environment variables

The "numpy fmod" issue

Our release notes contain the known issues for a specific release. We wanted to emphasize one specific known issue. Running the Orchid Python API example notebooks on Windows, the numpy package has presented an error like the following

RuntimeError: The current Numpy installation ('') fails to pass a sanity check due to a bug in the windows runtime. See this issue for more information: https://tinyurl.com/y3dm3h86s

Our work-around to this issue is to pin numpy to version 1.19.3 as recommended by Steven Wishnousky of Microsoft in the thread mentioned in the error. We are continuing to await a fix from Microsoft for this issue.