Working with high-Level tools weakens your math

I worked for 20 years with Numpy, Pandas and such solving high level problems.

But today I stumbled because of a simple linear equation:

a * x + b * y = c

In most case one will rewrite this to

y = m * x + d

and I did.

We all learned at university that this is not true for some special cases, but we usually don't care since the framework will handle the special cases.

While programing an optimized new numerical core this part of history hits me.

Deriving m=(Y1-Y2)/(x1-x2) when (x1-x2) == 0 is a problem.

So please learn from my mistake.