I have recently started to like using ipython to do a lot of things I used to use tools like Jupyter, Desmos, and Mathematica for. And through this, I have found a powerful new tool to add to my collection.
The Beauty Of the Command Line
I have always been someone who loves to do everything through the terminal. With the way I interact with my computer primarily being the keyboard(I try to refrain from using the mouse), the cli always ends up being the fastest way to do things.
For this reason, I loved programs like Neovim as my idea, as I am able to work at a speed I could only dream of a few years back.
But the biggest pitfall has always been interactive python development, with things like Jupyter Notebooks.
The pitfall of Text Buffers
Neovim, despite being easily one of the most customizable IDE’s there are, is still only a text buffer. And even with programs like Neovide which seek to fix this pitfall, there are some GUI-related things it cannot do easily.
The thing that this has hurt the most for me, has been Jupyter Notebooks. As someone who is very much in the world of AI and Deep Learning, Jupyter is the standard for combining research with production code. It allows us to blend markdown and annotations into our code, but also provides a more streamlined interpreter experience with python.
How iPython fits into this
iPython, while still only being a text buffer, allows me to keep this interpreter conversation style within the terminal.
But Ewan, how does this fix Jupyter Problem?
Doesn’t this mean it has the same problem that you had before, and you cant really blend your code the same way?
Well, yes, but not really. The simplicity of the ipython interface means I can just embed the command log as a python code block. This, in combination with the Code Runner extension, mean I can make powerfull scientific notes within my code projects.
[ins] In [1]: import random
[ins] In [2]: x = random.randint(1, 100)
[ins] In [3]: x
Out[3]: 63
[ins] In [4]: y
-----------------------------------------------------------------
NameError Traceback (most recent call last)
Cell In[4], line 1
----> 1 y
NameError: name 'y' is not defined
[ins] In [5]: Doing math with iPython
The thing that sealed the deal for me was one iPython extension: Copilot.
Copilot for iPython meant I could essentially have an Ai-assisted CLI(without wierd proprietary terminals, cough cough warp) in conjunction with a python interpreter.
Since you can easily run shell commands while still in an iPython session, this serves as a great way for me to quickly do programming tasks that I realistically don’t need to plan out - something like solving a calculus problem.
Now, I can solve an integral easily and quickly in the command line:
[ins] In [1]: # setup integrand
[nav] In [2]: integrand = sym.cos(2*x)**2 + sym.sin(4*x)**3
[ins] In [3]: result = sym.integrate(integrand, x)
[ins] In [4]: result
Out[4]: x/2 + sin(2*x)*cos(2*x)/4 + cos(4*x)**3/12 - cos(4*x)/4