Python 3 and multi-environment support, new state of the art package dependency solver, and over 450 packages now available free for all users
Enthought is pleased to announce the release of Canopy 2.1, a significant feature release that includes Python 3 and multi-environment support, a new state of the art package dependency solver, and access to over 450 pre-built and tested scientific and analytic Python packages completely free for all users. We highly recommend that all current Canopy users upgrade to this new release.
Ready to dive in? Download Canopy 2.1 here.
For those currently familiar with Canopy, in this blog we’ll review the major new features in this exciting milestone release, and for those of you looking for a tool to improve your workflow with Python, or perhaps new to Python from a language like MATLAB or R, we’ll take you through the key reasons that scientists, engineers, data scientists, and analysts use Canopy to enable their work in Python.
First, let’s talk about the latest and greatest in Canopy 2.1!
- Support for Python 3 user environments: Canopy can now be installed with a Python 3.5 user environment. Users can benefit from all the Canopy features already available for Python 2.7 (syntax checking, debugging, etc.) in the new Python 3 environments. Python 3.6 is also available (and will be the standard Python 3 in Canopy 2.2).
- All 450+ Python 2 and Python 3 packages are now completely free for all users
- Built in, state of the art dependency solver (EDM or Enthought Deployment Manager): the new EDM back end (which replaces the previous enpkg) provides additional features for robust package compatibility. EDM integrates a specialized dependency solver which automatically ensures you have a consistent package set after installation, removal, or upgrade of any packages.
- Environment bundles, which allow users to easily share environments directly with co-workers, or across various deployment solutions (such as the Enthought Deployment Server, continuous integration processes like Travis-CI and Appveyor, cloud solutions like AWS or Google Compute Engine, or deployment tools like Ansible or Docker). EDM environment bundles not only allow the user to replicate the set of installed dependencies but also support persistence for constraint modifiers, the list of manually installed packages, and the runtime version and implementation.
- Multi-environment support: with the addition of Python 3 environments and the new EDM back end, Canopy now also supports managing multiple Python environments from the user interface. You can easily switch between Python 2.7 and 3.5, or between multiple 2.7 or 3.5 environments. This is ideal especially for those migrating legacy code to Python 3, as it allows you to test as you transfer and also provides access to historical snapshots or libraries that aren’t yet available in Python 3.
Why Canopy is the Python platform of choice for scientists and engineers
Since 2001, Enthought has focused on making the scientific Python stack accessible and easy to use for both enterprises and individuals. For example, Enthought released the first scientific Python distribution in 2004, added robust and corporate support for NumPy on 64-bit Windows in 2011, and released Canopy 1.0 in 2013.
Since then, with its MATLAB-like experience, Canopy has enabled countless engineers, scientists and analysts to perform sophisticated analysis, build models, and create cutting-edge data science algorithms. Canopy’s all-in-one package distribution and analysis environment for Python has also been widely adopted in organizations who want to provide a single, unified platform that can be used by everyone from data analysts to software engineers.
Here are five of the top reasons that people choose Canopy as their tool for enabling data analysis, data modelling, and data visualization with Python:
1. Canopy provides a complete, self-contained installer that gets you up and running with Python and a library of scientific and analytic tools – fast
Canopy has been designed to provide a fast installation experience which not only installs the Canopy analysis environment but also the Python version of your choice (e.g. 2.7 or 3.5) and a core set of curated Python packages. The installation process can be executed in your home directory and does not require administrative privileges.
In just minutes, you’ll have a fully working Python environment with the primary tools for doing your work pre-installed: Jupyter, Matplotlib, NumPy and SciPy optimized with the latest MKL from Intel, Matplotlib, Scikit-learn, and Pandas, plus instant access to over 450 additional pre-built and tested scientific and analytic packages to customize your toolset.
No command line, no complex multi-stage setups! (although if you do prefer a flat, standalone command line interface for package and environment management, we offer that too via the EDM tool)
2. Access to a curated, quality assured set of packages managed through Canopy’s intuitive graphical package manager
The scientific Python ecosystem is gigantic and vibrant. Enthought is continuously updating its Enthought Python Distribution package set to provide the most recent “Enthought approved” versions of packages, with rigorous testing and quality assessment by our experts in the Python packaging ecosystem before release.
Our users can’t afford to take chances with the stability of their software and applications, and using Canopy as their gateway to the Python ecosystem helps take the risk out of the “wild west” of open source software. With more than 450 tested, pre-built and approved packages available in the Enthought Python Distribution, users can easily access both the most current stable version as well as historical versions of the libraries in the scientific Python stack.
Consistent with our focus on ease-of-use, Canopy provides a graphical package manager to easily search, install and remove packages from the user environment. You can also easily roll back to earlier versions of a package. The underlying EDM back end takes care of complex dependency management when installing, updating, and removing packages to ensure nothing breaks in the process.
3. Canopy is designed to be extensible for the enterprise
Canopy not only provides a consistent Python toolset for all 3 major operating systems and support for a wide variety of use cases (from data science to data analysis to modelling and even application development), but it is also extensible with other tools.
Canopy can easily be integrated with other software tools in use at enterprises, such with Excel via PyXLL or with LabVIEW from National Instruments using the Python Integration Toolkit for LabVIEW. The built-in Canopy Data Import Tool helps you automate your data ingestion steps and automatically import tabular data files such as CSVs into Pandas DataFrames.
But it doesn’t stop there. If an enterprise has Python embedded in another software application, Canopy can be directly connected to that application to provide coding and debugging capabilities. Canopy itself can even be customized or embedded to provide a sophisticated Python interface for your applications. Contact us to learn more about these options.
Finally, in addition to accessing the libraries in the Enthought Python Distribution from Canopy, users can use the same tools to share and deploy their own internal, private packages by adding the Enthought Deployment Server. The Enthought Deployment Server also allows enterprises to have a private, onsite copy of the full Enthought Python Distribution on their own approved servers and compliant with their existing security protocols.
5. Canopy’s straightforward analysis environment, specifically tailored to the needs and workflow of scientists, analysts, and engineers
Three integrated features of the Canopy analysis environment combine to create a powerful, yet streamlined platform: (1) a code editor, (2) an interactive graphical debugger with variable browser, and (3) an IPython window.
- Canopy’s code editor comes with everything required to write analysis code, but without the burden of advanced development environments like PyCharm or Microsoft Visual Studio (although, if needed, other IDE’s can be configured to use the Canopy Python environment). With syntax highlighting, Python code auto-completion, and error checking, users can quickly interact with Python code, write and execute existing code.
- Canopy’s interactive graphical debugger with variable browser helps you quickly find and fix code errors, understand and investigate code and data, and write new code more quickly.
- The integrated IPython window lets you quickly test code, experiment with ideas and see the results of code run directly from the editor. Canopy also includes pre-configured Jupyter Notebook access.
Finally, access to package documentation at your fingertips in Canopy is a great benefit to faster coding. Canopy not only integrates online documentation and examples for many of the most used packages for data visualization, numerical analysis, machine learning, and more, but also let you easily extract and execute code from that documentation to get started working with starter code quickly.
We’re very excited for this major release and all of the new capabilities that it will enable for both individuals and enterprises, and encourage you to download or update to the new Canopy 2.1 today.
Have feedback on your experience with Canopy?
We’d love to hear about it! Contact the product development team at firstname.lastname@example.org.