Now,I’m studying numby library to take deep learning specialization is courses contain videos about it or I must still learning it till I become perfect??

@Ahmed112 Numpy really is not much different than using a calculator. I feel one has to be comfortable with the math you want to achieve, before you use it.

But Numpy really has no ‘magic’.

In the Deep Learning Specialization they introduce you to numpy and give examples of how to use it, so you don’t need to be a subject matter expert in numpy before you start DLS. And as Anthony says, if you understand Linear Algebra and the various math operations involved there, numpy is just providing you with functions that perform the various types of math operations. So understanding the math is the gateway to everything else. For example, to compute dot products or transposes or the norm of a vector or matrix “there’s a function for that”.

Thank you so much

Thank you so much

Hey Paul, my Indian roommate in undergrad actually *was* a math major, and he often extoled to me the benefit of RPN on old HP calculators-- I never could quite wrap my mind around that, or its benefit. But luckily for our friend here, Numpy does not do that.

I’m still using my HP calculator from the late 1970’s.

RPN is like:

`<parameter> <parameter> <function> and you get a result.`

Algebraic is like:

`<parameter> <operator> <parameter> <equals> and you get a result.`

@TMosh I am sure it is not bad. He was (or became) my friend after all.

But he never explicated to me the advantages.

So, you are saying, if you can think in that way, it is just ‘less typing’ ?

I guess that makes sense.

I also do not like working in (micro)PIC assembly unless I have to.

I owned only HP calculators “back in the day”. Once you understand Reverse Polish Notation it is clearly the only way to go. The point is that you don’t need parentheses to control the “order of operations”: it is all explicit and easy to understand once you “grok” the concept. The key is to visualize the “stack” of operands that you have at any point in the process.

My point here was-- ‘jokes on me’, or to our friend @Ahmed112, Numpy is not that confusing/something that needs study. As a contrast, I wished to suggest something I always did find ‘really confusing’.

Sorry, I was being a bit too literal minded (not the first time for that of course) and obviously did not get why you were introducing the RPN topic.