How proficient do you have to be in C++ for an entry level quantitative analyst role?

Joined
9/19/15
Messages
100
Points
28
Many people say that C++ and Python are must for quantitative analyst roles these days. I am pretty good at Python and will start learning C++.

How proficient do I have to be in C++ for an entry level quantitative analyst role? I am not a CS major..


From this list, which books are must for beginners and are those books going to be sufficient for an entry level interview?
Thank you.
 
you don't need to be that proficient in C++ to be quant analyst. Get something like the Bruce Eckel books to start. Also make sure to read more about the recent improvements in C++: C++11, C++14 and C++17
 
you don't need to be that proficient in C++ to be quant analyst. Get something like the Bruce Eckel books to start. Also make sure to read more about the recent improvements in C++: C++11, C++14 and C++17


I chose to read (and hopefully to finish by summer) C++ How to Program by Harvey Deitel. Is this also going to be enough?
 
you don't need to be that proficient in C++ to be quant analyst. Get something like the Bruce Eckel books to start. Also make sure to read more about the recent improvements in C++: C++11, C++14 and C++17
C++17 is not shipping, yet, AFAIK. C++14 is just a few fixes for C++11.
 
maybe have a peep at C++ libraries like Boost and Quantlib in due time.

Smart pointers are essential.
 
Deitel's book (and probably Savitch's book as well) should be titled, "C++ for dullards."
I had a look at it once..

For me, I learned a lot from Bruce Eckel's book in 1988. He had some great examples that really kick-started the C++ library we created for CAD and engineering projects with AutoCAD.
 
So Deitel's book is not useful, you are saying?
A general remark on nearly all C++ books (including some of my my own) is that they are pre-C++11, which is not an issue I suppose. However, some books have an over-reliance on inheritance and class hierarchies while neglected more modern programming styles endangers 'future-proofness' . And as bigbadwolf seems to say: many examples (widgets, cats, dogs, cars, planes, miauw-miauw, woof-woof) are mind-numbing. My favourite is:

class IceCube: public Ice, public Cube {}; yikes..

C++ is moving to templates and the functional style in a big way IMO. For a glimpse, see

https://www.quantnet.com/advancedcpp/

So, 'going forward' I think it is best to adopt the C++11 style.
 
Last edited:
So Deitel's book is not useful, you are saying?

By itself, not of much use. Instead pick up a copy of Titus Beu's "Introduction to Numerical Programming." Amazon is selling it for $72, which is a steal. Work through it diligently. And keep a copy of Lippman (or something roughly equivalent) by your side for reference.
 
By itself, not of much use. Instead pick up a copy of Titus Beu's "Introduction to Numerical Programming." Amazon is selling it for $72, which is a steal. Work through it diligently. And keep a copy of Lippman (or something roughly equivalent) by your side for reference.

I am actually interested in risk quant, not front office quant. I think I should focus on Python almost exclusively..
 
A general remark on nearly all C++ books (including some of my my own) is that they are pre-C++11, which is not an issue I suppose. However, some books have an over-reliance on inheritance and class hierarchies while neglected more modern programming styles endangers 'future-proofness' . And as bigbadwolf seems to say: many examples (widgets, cats, dogs, cars, planes, miauw-miauw, woof-woof) are mind-numbing. My favourite is:

class IceCube: public Ice, public Cube {}; yikes..

C++ is moving to templates and the functional style in a big way IMO. For a glimpse, see

https://www.quantnet.com/advancedcpp/

So, 'going forward' I think it is best to adopt the C++11 style.

I am actually interested in risk quant, not front office quant. I think I should focus on Python almost exclusively..
 
From your first post:



You don't know what you're going on about and asking frivolous questions.

I was thinking of doing C++, until I talked to someone just a few hours ago. He said C++ is important in front office quant roles as 'speed' is very important. Speed is not important for risk quants and it is more worth to study python than C++ given my interest in risk quants..
 
Back
Top Bottom