Skip to main content

Visualizing python project dependencies

When you install python packages using pip, it remembers the dependency tree of installed packages. Together with Graphviz, pipdeptree can generate a nice visualization of python project dependencies.

What is even better is – It allows to save the output in a user friendly format such as jpeg, png or pdf.

How to use it on Windows 10 PC

  • Download Graphviz from this link
                http://www.graphviz.org/download/
  • Add below to PATH environment variable (Update the installed graphviz version if needed)     
       C:\Program Files (x86)\Graphviz2.38\bin 
       C:\Program Files (x86)\Graphviz2.38\bin\dot.exe
  • Close any open command prompt, restart it and navigate to your python project folder. e.g.,      
          cd C:\Projects\GitHub\Installer_Bootstrapper
  • Install ‘pipdeptree’ and ‘graphviz’ python libraries   
       pip install pipdeptree
       pip install graphviz
  • Now you can start testing 'pipdeptree'
    • View dependency tree of all installed packages 
                 pipdeptree
    • View dependency tree of a particular packages - django
          pipdeptree -p django
    • View reverse dependency tree 
                 pipdeptree -r

  • Now, let’s visualize it as jpeg image using GraphViz 
        pipdeptree --graph-output jpeg > out.jpeg
        
         or, render dependency graph as a pdf file 

        pipdeptree --graph-output pdf > out.pdf
        
This is how the generated image looked like 🙌🙌


Comments

Popular posts from this blog

Solving PyCharm bug: "Python helpers are not copied yet..."

I have been using PyCharm Professional to run and debug Python code from my Windows machine to a remote Linux device. There are some other tools available for the same purpose (e.g. Python Tool for Visual Studio 2017). But in my opinion, PyCharm Professional stands out among its counterparts as it comes with a "All Batteries Included" setup. Once you configure remote Python interpreter in PyCharm, then it works out of the box. However, today after upgrading to PyCharm Professional 2018.2.1, I could not run my Python script on remote device. The execution always failed with below error - "Error running 'hello': Python helpers are not copied yet to the remote host. Please wait until remote interpreter initialization finishes." To solve this issue, I had to remove the ".pycharm_helpers " folder from the remove device and then restart PyCharm so that the folder is re-created and files are copied again. Here are the steps with comma...

Threading tutorials from Microsoft

Do you know what’s lurking in the deepest, darkest parts of your thread pool? Take control of that monster and keep your applications responsive by becoming a threading expert with this free 5-part course. Advanced .NET Threading , Part 1: Thread fundamentals Advanced .NET Threading , Part 2: Compute-bound async operations Advanced .NET Threading , Part 3: I/O-bound async operations Advanced .NET Threading , Part 4: Thread synchronization primitives Advanced .NET Threading , Part 5: Thread synchronization locks

32 bit and 64 bit system files in Windows

The windows operating system stores both 32 bit and 64 bit system files under seperate folders.  Because 64 bit OS needs to be able to run 32 bit programs without confusing them with new paths, some paths are redirected to different folders depending on whether a 32 bit or 64 bit application looks at them. The same applies to some registry keys.  Basically, if you open explorer (which is a 64 bit program), you will see these folders: C:\Windows\System32 << your 64 bit system folder C:\Windows\SysWOW64 << your 32 bit system folder (Yes, the numbers appear to be exactly the other way round, looks like someone at Microsoft was trying to be funny!) However: When a 32 bit program, such as Atom, looks at your Windows folder, it will see these folders instead: C:\Windows\System32 << your 32 bit system folder ( !!! ) C:\Windows\Sysnative << your 64 bit system folder This ensures that under the path C:\Windows\System32 , all a...