Search results

These? https://www.bogotobogo.com/cplusplus/google_unit_test_gtest.php https://www.boost.org/doc/libs/1_75_0/libs/test/doc/html/index.html // I haven't used them as they are not in my job title at the moment :) But you probably need them in production environment. Certainly worth knowing a...

It's VS, so yes.

I forgot to say; Excel Driver only works with Visual C++. Not with dev-c++, for example.

I assume you tested it on the HelloWorld example?

Anyone using my driver on Excel 365? I/we get errors C:\Program Files\Microsoft Office\root\VFS\ProgramFilesCommonX86\Microsoft Shared\OFFICE16\MSO.DLL :128:4:error:stray '\374' in program C:\Program Files\Microsoft Office\root\VFS\ProgramFilesCommonX86\Microsoft Shared\OFFICE16\MSO.DLL...

Cannot stress how many "Math whiz" PhDs with strong PDE skills we have not hired because they don't have a strong SWE background Sounds logical. This happens when pure maths pde (theory only) have not learned to program. So, it would not be a surprise to me. It takes [7,20] years to really...

What's "SWE background"? What distinguishes "programmer" from "developer"? It may be an idiosyncratic and have several definitions on who you ask. I would say most quants of any flavour should be able to test their ideas (in a computer).

In an academic context, the Association for Computing Machinery defines IT as "undergraduate degree programs that prepare students to meet the computer technology needs of business, government, healthcare, schools, and other kinds of organizations .... IT specialists assume responsibility for...

Let me qualify. 'IT' is a term that emerged during the outsourcing bonanza of the 1990s. My bad. I should have known better not to use that word because it does not mean what many people think it means. Let me rephrase: how many software companies that originate, develop and market software...

Can you name one Russian IT company? Historically a challenge is when software has to written to fit into 16 bit machines. And a soclal price was paid if they discovered you were a prodigy..

One way (and not a bad way) to learn maths is to reverse engineer an algorithm by programming it in C++ and work backwards to the original maths formulation. I do that all the time these days.

Not knowing both SDE and PDE could be a drawback. Still, most models are built around these pillars I reckon. Of course, the lack of can be compensated by other talents like programming ability, for example.

and that's fine. It stops it seems at 1-factor diffusion?

Thanks. I don't want to come across as arrogant but the MIT contents are 2nd (1st?) year freshman content. The heat equation is far too simple. A lot of quants go through life without understanding the subtleties of convection-diffusion-reaction PDEs. I do all this stuff as well but more for...

VBA was very popular. I am not sure in anno 2021 if it is worth doing a course. Maybe better to learn Python these days.

BTW I am also the originator of C++ Programming for Financial Engineering Advanced C++ and Modern Design // See "Financial Instrument Pricing using C++, 2nd edition" (Wiley) 2018 by Daniel J. Duffy

Regarding ODE/PDE, I am the originator of this course (recognised by UCB) https://www.datasim.nl/onlinecourses/97/distance-learning-ordinary-and-partial-differential-equations It is probably the only one geared up to finance PDE (Black Scholes) and FDM. Student discount, contact me. As...

Maybe a maths course at a community college to get to a 6-7 before the next step? You can compare contents with my online maths courses here https://www.datasim.nl/onlinecourses QN C++ course does no harm at all.

what's your current maths level?

“Young man, in mathematics you don't understand things. You just get used to them.” ― John von Neumann

n/a

Can you program in C/C++/Python? And I reckon some more training in maths is probably needed, depending on what "Mathematics" you did..

C++23 (in the future) https://en.wikipedia.org/wiki/C%2B%2B23 I would say; no big new features, rather fixes and extra thingies added to C++20.

Here's a very simple runnable example to show how Concepts work. // Test101Concepts.cpp // // Simplest example of a system. Context diagram consists only // of Input and Output systems. // // We use C++20 Concepts to replace policy-based design // // Composition // // V2: design using C++20...

How do I know? I have a phd. Well done, matey. Dime a dozen. Freshly minted PhDs tend to have this arrogance. Industry experience will cure you of that soon enough :) BTW I have a PhD in Mathematics (PDE, FDM) (1980). here's some of my connections in academia and industry in EU, UK and USA...

yes and no. Research is 1. to get a PhD 2. after PhD (then become administator, delegate research to your PostDoc team).

... SDE sde(myOption); SDEAbstract<SDE, double> sud(sde); https://onlinelibrary.wiley.com/doi/epdf/10.1002/wilm.10606 That article used C++11 functions to model stochastic differential equations (SDE) which resolves efficiency and maintenance problems of class hierarchies. Now that C++20...

"Regarding C++98/03, I am glad to have learned it instead of the modern revisions." ...

As I wrote in a previous message (we are going in circles), "The constant need of knowledge update and of learning new tools is the main reason of burnout in software development.". Not at all. It just feels like that at the moment. You'll find out in due course after you land your first job...

That's for me to know and you to find out. Besides, it's all in the preface I posted here last Friday. Of course, you read it. You like supersonic conservation laws, so you do, what's special? Concorde V2?

e.g. Compilers have always lagged behind language specifications, that's also a fact. If Visual Studio 16.10 supports C++20 now, it's an exception to this rule that does not make it less true. This is a remark that not even wrong. It's a variation of the True Scotsman Fallacy. In that message...

I've finished _everything_. So it's a matter to get it into the forum. Very soon indeed. I would say I have done 80% at least of C++17/C++20. and C++20 will be at the end of the course, so you can start.

1) You mean, 1 factor? this approach is limited. It's a 1-trick pony. Besides, it has an analytical solution, so why bother? 2) no, because they don't crop up in finance. Lax-Friedrichs is only for advection PDE and is not unconditionally stable. Lax-Wendroff is mentioned for Asian option PDE...

Please try to investigate before posting half-truths. It just wastes so much time.

It's called progress. Strange reply and incorrect. If you had followed the C++ committee down the years you would know why. C++11 and Boost C++ use this stuff .. it's called Template Metaprogramming. Those GOF patterns are 30 years old. I use them since 1992,. They were OK when C++ was a baby...

C++20 is shipping; It supports Concepts and it will subsume traditional Design Patterns. For example, here is the Bridge pattern, it' the best way to do it. 42 This is coming on Advanced C++ course. // Interface contract specification template<typename T, typename Data> concept IDiffusion =...

A bit touchy? I am giving my opinion/experiences. I have > 200 maths books in my library. And hundreds of articles dating from the 70s. And my own research notes. And I work everything out with pencil and paper before surfing, it at all. Sometimes I search, but it throws up existing sources...

Surfing the Web is what you do in free time. Besides, 90% of the information is useless.

Surfing the web, LOL.

A short history of PDE methods as applied in finance: It is better to have 5 solutions to 1 problem that 1 solution for 5 problems (each problem has its own peculiarities and recipes may break down and then Plan B is needed). A big problem IMO is that the same old FDM models are being used on...

There are 4-5 other ways to compute boundary conditions. It's all a bit subtler than a payoff, Chapter 23 of my new book. It was a flash in the pan : mprof was last seen in 2014...

BS PDE does not have/need boundary condition, but the FDM does need numerical BC. The above reasoning is outdated and crude, I must say. Maybe ask your prof then. Or a PDE quant. Send him/her my TOC and preface. I am available 24/7 for feedback. Or ask a PDE quant to discover what the real...

Au contraire; it's a 2021 PDE/FDM book! ' Did you miss the two Word files I posted here. Forget transforming to a heat equation, is not even wrong. It's a false sense of security. It's dumbing down. Anyway, linear two-factor PDEs are of convection-diffusion-reaction type with mixed derivative...

Pde and finite differences are to be avoided for multi-asset products as the multidimensionality of the problem hits hard on computational efficiency PDE up to dimension 3 are doable. They weren't built for higher dimensions. It's a no-brainer to use MC in those cases. Some of the ML community...

As for PDE, it's just a numerical physical exercise to get numerical value of a derivative given a model. You know what most of C++ developers can just copy and paste free source codes they found somewhere on the web and implement it and wah lah...they get the price of a derivative by the end of...

You might be interested in knowing that a PDE/FDM book is on the way.

I would say it is better to learn both PDE and SDE methods. They are two sides of the same coin. I fact there is a 1:1 correspondence as Einstein showed in 1905, See also the book by Friedman. So, both are very handy...

I used Bruce Eckel's book way back. It might be an idea to check his more recent(?) books.

Most of the C++ writers have given up the chase or have sadly passed. A remark: learn C syntax before C++, if not it is NOT optimal not at all. typedef is very limited. The code on the site looks neatly laid out. A bit in the Quantlib style. I have Datasim Press with Boost C++ books. In this...

One more thing .. my books are reasonably priced and all the time and effort is from my own pocket .. I don't have crowdfunders not subsidies in that sense. I think academics should also give something back to the students. That's my opinion,.

I don't have the book. From the TOC I don't see much C++11 buzzwords. Just sayin'. BTW my basic C++ doesn't cover C++11, I have a different focus, having come from an industrial background where we focus on well-designed and maintainable applications. Each to his own, I suppose. The QN/Baruch...

And this book by Luigi Ballabio on Quantlib library has lots of finance. Someone who has done QN C++ will have no difficulty using this library. https://www.amazon.com/Implementing-QuantLib-Quantitative-finance-architecture/dp/B08KHSZK86

Here is C++20 preview which will be in QN/Baruch C++ update very soon. https://www.datasim.nl/application/files/6716/1131/6643/Global_Overview_Cplusplus17_and_Cplusplus20.mp4

I looked at TOC. Seems to be well-structured, pedagogically. It is probably pre-C++11 so it won't age well. Does author provide source code? Hefty price. I wold say the next generation C++ apps should be C++20.

5

The only modern C++ books for finance are the ones that I have written. If C++ is new to you I recommend QN C++ (I am the originator and APalley is TA). Can't go wrong. Best course on the planet:cool::ninja::coffee:

methods versus analysis.

really? What about Paris VI, it's highly mathematical https://le-blog-finance.com/france/finance-le-master-el-karoui-forme-un-quant-sur-trois-en-france/ again, mathematical engineering as a term only exist in the politecno , no on else uses it AFAIK.

Have u searched on QN?? https://quantnet.com/threads/french-mfe-programs.6538/

hmmm There's probably not a university on the planet that doesn't offer such a course.. On my coffee break I found (caveat: I am not an expert in schools). 1. Best schools for Finance in Europe – QS Rankings 2021 London School of Economics and Political Science (LSE), the UK University of...

Difficult; The only one I know in this area I vaguely know is Bocconi. How did you arrive at this list in the fist place?

Bocconi (typo)

Indeed. In many cases the best way to understand a model is to work out all the steps by hand in simplest case with pencil and paper, e.g. 2X2 case. Then automate the process.

Anyone out there used Visicalc on Apple II https://en.wikipedia.org/wiki/VisiCalc It represented a revolution in personal computing. It was popular with e.g. cost engineers. "VisiCalc isn't as easy to use as prepackaged home accounting programs, because you're required to design both the...

That's probably not a goal. Compare it to _using_ a Pyton library, you don't need know the internals. Most plumbers have no idea what the Navier Stokes PDE is.

Congratulations; your future is in your own hands.

The nice thing about Bronson is that it's got good examples that can be done with pencil and paper. Once that is done means the step to coding is easy and better than trial-and-error programming.

Another scenario is solving the Portfolio optimisation problem by using the existing libraries in Julia. Of course, it is possible to write your own code but why bother unless you want to become a numerical analyst?

Do you mean Nocedal and Wright? mathematically advanced and too theoretical for the current project. It is black belt stuff. Bronson "Operations Research" (Schaum) is super! A bit of free/non-binding advice.. I use Nocedal _And_ Nocedal at different times for different reasons. Python has...

That's 3 questions 1. optimisation 2. Julia 3. Finance Learn all 3 at once?? I would focus on least squares in C/C++. Then goto next level. https://en.wikipedia.org/wiki/Linear_least_squares

You can see this by noting how many top quants specialized in pure math topics such as number theory. It this the reason why many quants don't know PDE/FDM? Computational finance and 'pure' maths are miles apart.

What's in a name? I worked in Maths _and_ Engineering; Mathematical Engineering is a meaningless name IMO, unless someone DEFINES it! Engineering Mathematics does exist https://en.wikipedia.org/wiki/Engineering_mathematics

Looks as if it is focused on engineering, architecture, chemistry, aeronautics. They might have a finance 'enclave'?? Bocconi in Milan is more prestigious.

Hard to say. It's a description. Not a detailed table of contents.

TBD, scientific computing can mean anything. Do you mean 'programming'?

Not necessarily. First do some background research ;)

These *binary* questions are drifting from the initial question. Too general. For PDE/FDM C++ is good, However, a great way to learn FDM is to write prototypes in Python.

Long story short: PDE libraries in Python has NO point. Python wasn't built for that. Libraries are built by C++ programmers and used by Python developers.

Maths theses need time to incubate, indeed. Exx. generalise CSM to n dimensions and use it instead of Automatic Differentiation (AD) in ML, especially in gradient descent etc.

You've already told us all that; you're hogging the forum.

I hope you do my PDE/FDM/C++ courses!

Depends on your mathematical sophistication. Hull is practical.

2. I'd like to do a PDE-based project. Any tips on where I should look for project ideas? Project idea do for UnCertain Vol,Heston,SABR. Here is a very good MSc PDE finance thesis from 2015, supervised by DD. Quasar; see also www.datasim.nl for theses 2019/2020.

I do PDE/FDM finance project as part of my online courses. If you wait a few months you can see them in my forthcoming book. And many are programmed in my 2018 book Financial Instrument Pricing in C++ (Wiley). One-stop shop :) Logan's book is great if you are studying biolgy. A bridge too far...

Buy Hull and Joshi.

Good :-)

C++ treats variables as criminals, Python treats them as consenting adults.

This https://en.wikipedia.org/wiki/Variable_(computer_science) and then in Python versus C/C++.

Python is A big difference between C and Python is in the area of variables A-Z.

No idea.

?

Out of curiosity, which programming langauage(s) do you know?

At the end of the day, programming entails 2-3 fundamental activities 1. Variables 2. Functions 3. Creating algorithms that use 1 and 2. Maybe C is too rough just yet, but a good way might be VBA in Excel. file:///C:/Users/DATASI~1/AppData/Local/Temp/VBAPrimer.pdf Nice thing about Excel is...

Sorry to hear that.

Did you try programming them or just reading? The point is you get support from TAs which cannot be encapsulated in a book. Hands-on is what matters. Like learning martial arts, you need trainer.

The Quantnet C++ course assumes no prior knowledge of C++, A wise decision was to do a couple C modules. In fact, quite a few students had never programmed before. https://quantnet.com/media/categories/level-1-videos.2/

Mostly PDE/FDM pricing libraries in C++ for speed and versatility. I give all these online courses.

The C++ part is outdated; it is early 90s. It is still useful to know but C++ has moved on.

Some useful do and don't when designing applications in C++. 1. Multiple inheritance was not a good idea. C++ does not support interface keyword. The AI community used MI a lot in the 90s. 2. Not evertyhing is an object. Behavioural commonality is not structural commonality.This messes up class...

C++ is one of the few serious programming languages that helps you develop into a competent developer. The various skills that you learn cannot readily be developed in other languages because many of them treat you like a child, or hide important details that you will never get access to or even...

I have no real idea. Like 2nd hand cars, they can be cheap and cheerful, and pretty much useless. QN C++ is great value for money.

This C++ book by the late Dr. Joshi is severely outdated. The best way to learn C++ is to take Quantnet C++ online course. One learns C++ by doing, not reading. I have several books (lost count how many). See also...

Fortran is great, but very very niche. Python is popular. C++ is a skill. background is in applied maths, physics and numerical analysis. I know a lot about solving differential equations and high-performance compute I have a new PDE/FDM book (Wiley) coming out in a few months, so right up...

How's your C++?

Bla, bla bla. Heard of smart pointers? Try this There are two kinds of programmers 1. those who do C++ 2. those who talk about C++ (ad nauseam) Anyways, you did not address my queries on Rust.

KKOP rocks. It also has linear algebra and more.

Sounds too good to be true. Is it a one-trick pony? Rust is #24 in the top 100 Billboard charts. https://www.tiobe.com/tiobe-index/ Sounds like waffle, your post. No offence intended.

R is a tool with lots of useful packages, C++ is a programming language. They serve different purposes.

Why not ask your IT department??

I am not an employer but my feeling is if you know this modern stuff it proves you are flexible which employers like.. Especially new startups, maybe. Good luck! I assume you have a supervisor whom you can chat with for some topics as well.

aka intuition, nothing wrong with that. Of course, knowing how to solve problems beforehand is better.

I feel for all the students who have to learn this stuff. I had at least 3 courses as undergraduate related to measure and it was super abstract. What SDE books need IMO is more applications and numerics on top of the theory.

I have found Green's function to be not so useful. On the one hand it is PDE theory (existence of a solution) and on the other hand it produces a solution as an integral. It is limited to simpler problems. Still. it gives great insights into PDEs. The book by Stakgold (1998) is worth a read...

Like Johnny von Neumann said, maths is about learning the symbols. My way is ODE -> bit of ODE numerics -> PDE theory -> Black Scholes -> BS numerics ->(optional) small BS project in C++ or Python. Ready to go. Trajectory takes 2-3 months for student entering MFE/MSc. Euler, Gauss, Lagrange...

Off tangent, but I feel like this forum over emphasizes PDEs. I got into the fin econ program at Columbia which goes through their PhD curriculum, MIT, and UChicago + got interviews at some other places and I’ve never taken a PDE course. PDE knowledge is taken from other backgrounds. An econ...

I never took PDE in undergrad and got a C+ in ODE, but I was an applied maths major, I can kind of understand that pure maths don't do PDE (and at the same time not) but for applied maths PDE theory + numerical PDE would be essential. Maybe I'm missing something.

You're welcome. On www.datasim.nl (BLOGS) there are 3 recent and super-modern theses to be found.

Are the results 'interpretible'? https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3829947 Something auditors and regulators like.

Some are on the sites of several US universities. Which is a sort of accreditation I suppose. And students learn a lot. Student discount also.

I offer most maths courses https://www.datasim.nl/onlinecourses https://www.datasim.nl/testimonials

The beginning of wisdom for a programmer is to recognize the difference between getting his program to work and getting it right. A program which does not work is undoubtedly wrong; but a program which does work is not necessarily right. It may still be wrong because it is hard to understand; or...

At first glance, this looks good and practical. Possibly a bit more direct coding with your own template code does no harm. One question relates to how hands-on it is 1) understand the concepts, 2) program them in C++. For 2), the C++ courses have lots of exercises and brilliant support from the...

Are you shifting the goalposts?? Now it's Python v C++? From a business manager perspective, is Rust future-proof? In August 2020, Mozilla laid off 250 of its 1,000 employees worldwide as part of a corporate restructuring caused by the long-term impact of the COVID-19 pandemic.[43][44] Among...

A bit like the videotape format war. https://en.wikipedia.org/wiki/Videotape_format_war The main determining factor between Betamax and VHS was the cost of the recorders and recording time. Betamax is, in theory, a superior recording format over VHS due to resolution (250 lines vs. 240...

Penny-wise, pound foolish. It's strange that if a motivated working quant (who already works with production C++ libraries, say) wants to improve C++ skills, more time than not they have to pay out of their own pocket.

Dissertation and Miniprojects - These were very time consuming - you need to do all of these during Trinity term. That's around 60-80 pages of stuff you are required to write in 8 weeks or so. On the other hand, these projects were a good way to reinforce material learned in the first two...

C++ Programming Courses - You learn the basics and then learn how to use and extend the lecturer's own library. This is inadequate preparation for quant interviews and actual quant work. You will need to study some Comp. Sci (e.g. sorting algorithms) yourself. Baruch has resolved this problem...

Good question. I have been doing design in C++ since the 90s. You might find some useful background here. https://quantnet.com/threads/blog-articles-on-c-11-and-computational-finance-by-daniel-j-duffy.32237/ I am probably the only one on the planet who writes C++ books for finance. I prefer...

“An expert is a person who has made all the mistakes that can be made in a very narrow field.” ― Niels Bohr

Well, Barbara Streisand likes NYC

My ODE/PDE course is recommended by several universities. www.datasim.nl

It's what I've been saying since 1989!! That's when I started Datasim centred around C++, OOP and applications. Justifiably known as one of the most difficult coding languages to master, I know the way C++ feels And all of the hearts that it steals C++ tells me it's easy When you do it right

At university we learned the axioms of Natural Numbers -> Integers -> Rationals -> Dedekind cuts -> Real numbers step-by-step (e.,g. each natural number has a successor). It was like learning Theology. Before the IEEE 754 era. Now I can use these numbers in my sleep. // Now, here is maths...

well, good to see that at least 1 math joke exists, but it is not unique.

Well, we don't know until you try. It's like throwing a judoka into an MMA ring. Most pure maths is not constructivist https://en.wikipedia.org/wiki/Constructivism_(philosophy_of_mathematics) My favourite is Q: what is 2 + 3? A: the same as 3 + 2 Q: why?(?) A: because addition is commutative.

I had a look at the syllabus, good content https://www.maths.cam.ac.uk/postgrad/part-iii/part-iii-guide-courses It leans more to theory than applied. Just saying. Do you learn Navier-Stokes PDE?

It depends. Maths is not a magic elixir. I did all that pure maths stuff, but much of it is not so useful for finance. Being good at X means you are good at X. Analogy; a judoka is a useless karateka and vice versa.

The aim of part iii is to prepare for maths research. This is not exactly what is needed in finance. If you do pure maths, you miss out on {numerical, applied, computational} maths. Just sayin' And you have to be good at programming.

What about Luxembourg?

https://www.efinancialcareers.com/news/2021/03/c-jobs-banks Even where systems have been built in Python, a core of code is often built in C++. This is the case, for example, for JPMorgan's Athena pricing engine, and explains why the bank is currently looking for a C++ engineer for its core...

Answer: because not enough US citizens are! It's a more general trend that started in the 80s in Europe and I suppose USA. Many European graduates wanted to become managers. Countries like China, Morocco, France and etc. love maths. It's like decline of the West in a sense, Germany seems to...

Normally I have 64 bit across the boards. My Rule is Let;s say Boost has a new version 1.75. Use is after about [6,9] months. Use stable 1.72. When 1.80 comes along, then use 1.75. No time to be a beta/acceptance tester for Boost.

working 9-5 in production and blue sky research in the evening doesn't work, in general.

It doesn't work like that.

"Dr. TopQuantWannabe" it's timeless.

"A PhD is more of a character test than intelligence test" Dr. Sinead Kane (visually impaired ultramarathon athlete, double PhD doctorate, and double Guinness World Record holder, Dr Sinead Kane.).

Don't be lazy, be clever https://docs.scipy.org/doc/scipy/reference/generated/scipy.integrate.solve_bvp.html I reckon the Jacobian is taken care of in the solver? (default, J is done for you by fd). God luck; let us know how you get on. y'' + k * exp(y) = 0 y(0) = y(1) = 0 for k = 1.

you are doing sumpin' wrong.. You have to tell him WHERE to find lib (additional directories under Linker). If you have used Boost before .. or go back to 1.72 unless 1.75 is really necessary ;) check project 64 bit with you x32.,lb

On a follow-on remark (looking in the crystal balls), mainstream ML is grounded in linear algebra due to its (chance?) influences in the 80s in the seminal paper by Rumelhart, Hniton and Williams (at least that's the way it seems to me). Then everyone latched onto gradient descent? Linear...

This question can be broken down into several use cases. 1. Learning the syntax of linear algebra and its relationship to matrix theory. Gotta get used to symbols, notation and jargon ASAP. 2. Some numerical linear algebra. 3. Stuff like gradient, Jacobian, Hessian 4. Some background on...

Some MFE do ask for pdes. ode w/o pde is possible but why not do pdes, especially when it is for Black Scholes?

This is weird. Still doesn't answer my question. BTW if your pet goes back into the wild then it is not an instance of Pet? Here's a waccky example from the mid 90s. class ICE {}; class CUBE {}; class IceCube: ICE, CUBE { };

The code uses rand() .. meh Will rand() ever be deprecated? - C++ Forum rand() is not recommended for serious random-number generation needs, like cryptography. It is recommended to use C++11's random number generation facilities to replace rand(). (since C++11)

Don't get me started!. e.g. any one who writes a Pet class is for me a warning they have not used C++ in any kind of real app. Coding style is 90s as well.

This is awful code. class Pet .. that's a joke. The kind of stuff OOP did in the early 90s. In QN we can create classes for PDE, Monte Carlo etc.. And your TA is the 1-in-a-million Avi.

Who is "Anand"? I had a look at the courses' outline 'bullets'. The topics look standard if not somewhat outdated (90s feel) and missing stuff. 1. any samples of code? 2. testimonials?

There's no comparison between any C++ course and QN C++. At least I don't know of any. a guess: a semester of C++ at uni is about [20, 40]% of QN C++. To give a course in C++ you have to be an expert on C++, You can't do it on the side, like a weekend hobby.

And complete metric spaces. Cauchy sequences. And more Constructive Analysis ... prove existence and uniqueness of a solution by actually constructing it. Good example is Banach's fixed point theorem in metric spaces .. it can even be done in C++...

Mathematics is for ever. An example: I studied and researched in Functional Analysis and Hilbert spaces almost 50 years ago, now it is a new wave in Machine Learning.

LOL. You are projecting. No skin of my nose. If you have more time than money then you can scavenge internet ad infinitum to find online C++ stuff and it won't cost a penny.

yes and no. More important in my opinion is to understand the symbols an to applications. It should not descend into dry theorem proving alone. Metric spaces are a small part. See my hands on + exercises approach here. And is geared to finance...

Some of my issues with Gradient Descent @Cuchullain, for me, Gradient Descent is a swiss-knife methods. Always produce results, but can be stuck in local minima. Local minima, if it is lucky, That's the least of your worries. GD has a whole lot of issues: Off the top of my head 0. Inside GD...

It's from the swinging sixties :) We are now in the roaring twenties!

And Easter eggs.

https://c8.alamy.com/comp/CFF672/the-mary-quant-shop-in-duke-of-york-square-chelsea-CFF672.jpg

Caveat maintaining legacy code can be challenging. “A Big Ball of Mud is a haphazardly structured, sprawling, sloppy, duct-tape-and-baling-wire, spaghetti-code jungle. These systems show unmistakable signs of unregulated growth, and repeated, expedient repair. Information is shared promiscuously...

I have all the code ready. I just need to do the videos. Sorry for the slipped schedule (had a new pde/fdm deadline..) So, soon 2-3 weeks I reckon Here is video #1 for kick off https://www.datasim.nl/blogs/30/videos

welcome on QN

RKHS methods for ML Video, Jean-Marc Mercier and Daniel Duffy

"The mind is willing but the body is weak" ?? One hit wonder?

Algebraic geometry seems to have acquired the reputation of being esoteric, exclusive, and very abstract, with adherents who are secretly plotting to take over all the rest of mathematics. In one respect this last point is accurate. David Mumford

A number are on the MFE web sites that recommend them, in particular ODE/PDE and Numericals. And they are more in-depth than anything I've seen (and students say so too). And I am the originator of both QN/Baruch C++ courses.

https://www.datasim.nl/onlinecourses https://www.datasim.nl/testimonials

Some background on this talk Hilbert Space Kernel Methods for Machine Learning: Background and Foundations Daniel J. Duffy, dduffy@datasim.nl Jean-Marc Mercier, jean-marc.mercier@mpg-partners.com Datasim Education BV www.datasim.nl MPG-Partners Abstract Daniel will, in the...

Talk by Jean-Marc Mercier and Daniel J. Duffy on Hilbert Space methods for Machine Learning https://quspeakerseries24.splashthat.com/

I'm not sure if this is the right question.

I was referring to persuading pure mathematicians to show more inferest in mathematical physics! Academics tend to live in silos in the sense that they focus on very narrow subject areas. As I mentioned, this is not my area. And I don't see many real applications. And I have so many other...

Too late. It ended with Poincare. It's like trying to teach judo to a boxer.

My Topology prof was a PhD student of Hodge. Now, homotopy is useful in solving nonlinear systems of equation F(X) = 0 by embedding it in a larger H(x,t) = 0 and managed to use it later in industry.

I have them on my bookshelf. The topics are pretty dull. Where is this stuff used?

Your words, not mine. Those topics I did by the end of my 2nd year and most was 1st year stuff for us. And then I did 2 more years of more advanced topics. For the record in years 3 and 4 I did functional analysis, several complex variables, numerical analysis X 2, statistics, Topology of...

Another remark: after 4 years of undergrad you should be ready to embark on baby research instead of yet 1 1/2 years of cramming other's work. It's like a straightjacket.

In the case of bbw, I think he is way behind the curve. A little knowledge is a dangerous thing. Around 1900 mathematics fractured into pure maths and physics and both were damaged. I have the books by Wallace and Fulton on Algebraic Topology .. lots of pictures and flow diagrams but little...

It's the way ye tell it. Have you studied the PDE works of Lagrange, Laplace, Euler, Hamilton, Navier-Stokes, Poincare? PDE doesn't mean what you think it means.

Are we entering a 'no true Scotsman' discussion? In fairness, they are paid to do PDEs, not homeomorphisms.

I Purest of pure maths. My son is drummer in a metal band; the base guitarist + composer is doing a PhD in algebraic topology. As soon as I mention PDE etc. he freezes. Two mathematicians separated by a common language.. The two other guys are CS, so there is some rapport :devil:

In any case, I would lean C++, ideally at QN. C++ is great in combination with maths. Maybe your MFE school will do it come the time? BTW here is hard-core Numerical Analysis sub-topics https://en.wikipedia.org/wiki/Category:Numerical_analysis

In very general terms, most students tend to be good at/like pure math XOR applied/numerical maths. A bit one-sided and somewhat myopic.I had the luck of doing all 3 (as well as theoretical physics) as undergrads and we hardly ever used the words pure/applied in the context that they are used...

The best way to learn proper programming is QN C++ course. It's got 30 years of industrial (not uni) OOP and software know-how built into it.

This is a cookbook on how to use certain Python libraries. yes? Personally. I use the reference manual. Actually, many Python books are like this: 50% output data (just look at page 352-353 of McKinney, unused space), 30% plots, 15% methods and function summary, 0.1% maths, 0% software design...

In software, the ISO 9126 quality issues https://en.wikipedia.org/wiki/ISO/IEC_9126 And is your s/w a 1. throwaway prototype proof of concept 2. 1-man production systems 3. what's your budget and experience? And a million other questions. In general, my guess is pipeline (pipes and filters).

A PhD is a kind of degree you do because you _really_ want to do it. If that spark is missing then life is more difficult. https://jcs.biologists.org/content/121/11/1771 Getting the right answers is not the same as asking the right questions. They are separate skillsets.

And courses are 1. Half don't do homework. 2. Always 1 guy who's out to ruin things. But a nice way to meet people. I learned basic Dutch in the beginning by listening to songs and watching TV commercials.

It's context-sensitive and it depends on what the requirements are. There are different levels, An extreme case is you _think_ in your 2nd language and not in your mother tongue. I knew a Russian maths prof who after 6 months gave lectures in Dutch on "Generalised Functions and PDEs". Not bad...

Does no harm. If you are good in languages these classes will bore you. Join a (sports) club, mingle and get to know people.

Call it what you want, In some European countries you won't be taken seriously if you don't speak he language reasonably correctly.

So, where to learn German?

It's an example of linguistic ability, the test is does the student understand grammar in all its forms. People learning foreign languages tend not to get beyond present tense. "I don't smoke, but if I do smoke, I smoke Marlboro"

Are the lectures in Germany through English?