|
Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks. |
Looks like my class was only using ruby for project 1. Apparently the next 3 projects are now functional programming in OCaml. The language seems a little odd, since every language has been fairly similar up to this point. Except C I guess, but even that wasn't as much of an oddball as ocaml.
Anything crucial I should know? I know that everything is immutable in ocaml, which I guess is kind of the point with functional programming?
So no reassigning values within the same scope then? Any other tips from someone who had to suddenly switch to functional programming from imperative?
In other news it looks like I will be doing some research this semester! Going to be helping someone from NASA develop python scripts to predict changes to polar ice. So, climate change stuff. Sounds worthwhile and is relevant to my interests.
|
On September 15 2017 22:45 travis wrote:
Looks like my class was only using ruby for project 1. Apparently the next 3 projects are now functional programming in OCaml. The language seems a little odd, since every language has been fairly similar up to this point. Except C I guess, but even that wasn't as much of an oddball as ocaml.
Anything crucial I should know? I know that everything is immutable in ocaml, which I guess is kind of the point with functional programming?
So no reassigning values within the same scope then? Any other tips from someone who had to suddenly switch to functional programming from imperative?
In other news it looks like I will be doing some research this semester! Going to be helping someone from NASA develop python scripts to predict changes to polar ice. So, climate change stuff. Sounds worthwhile and is relevant to my interests.
That project seems pretty awesome! Too bad you're going to be incredibly depressed afterwards when you figure out how quickly the ice is melting... 
|
I'd love to work on such project.
And regarding your classes: why the hell would they choose OCaml to teach you functional programming instead of something that's more widely used in the real world? There's Haskell, Erlang, Clojure... Technically, you could do functional programming in Ruby and Python too, you just wouldn't be using some of the language's features.
|
I think we use it because some of the school's sponsors push it
anyways I already ran into a snag, and I am reading a lot about it but it's just so confusing to figure this out
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> true
This is fine. But not what I want. I want to see if h = t
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> (h = t)
not ok...
"Error: This expression has type 'a list
but an expression was expected of type 'a
The type variable 'a occurs inside 'a list"
ok so.. it evaluates to bool but I can't use it because it's not the same type as everything else? well then how the hell am I supposed to do this in a reasonable way?
|
|
|
The error says that the expression h=t is not valid. h is an element and t is a list.
If the list is 1,2,3,4
h=1
t=2,3,4
|
On September 16 2017 10:22 travis wrote:I think we use it because some of the school's sponsors push it anyways I already ran into a snag, and I am reading a lot about it but it's just so confusing to figure this out
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> true
This is fine. But not what I want. I want to see if h = t
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> (h = t)
not ok... "Error: This expression has type 'a list but an expression was expected of type 'a The type variable 'a occurs inside 'a list" ok so.. it evaluates to bool but I can't use it because it's not the same type as everything else? well then how the hell am I supposed to do this in a reasonable way?
I haven't actually used OCaml or really any real functional language, so take it with a grain of salt. But from what I know, you're matching the input against a couple of patterns here:
[] -> lst is an empty list
h::[] -> lst has exactly one element h followed by an empty list
h::t -> lst has exactly one element h followed by a list t (tail) which could be empty if it wasn't for the the previous pattern
When deconstructing a list like that, you always have a single element as the head and the original list minus the head as the tail. The tail is always a list.
Now you obviously can't compare single elements to entire lists, OCaml has proper typing as far as I know.
What exactly do you want? That the tail t is also exactly one element and that element is equal to h? That the very last element in the tail is equal to h? Something else?
I'm not sure what OCaml offers you here, but for the very last element, I'd expect to have to use some existing function or recursion. For 1st = 2nd in lst, either you can match something like h1:h2::t (I have no idea what the syntax would be, intention is to match "2 heads" and the tail), or you use another pattern matching on t inside the 3rd pattern.
I think it's ok to use OCaml at university. It's a decent language from what I hear, at least for academic purposes. The important point is that you learn how functional programming works, and that's fairly independant from the actual language. If you want to be capable of using a language productively (at a job etc), you probably will have to put in way more time than what you'd need just for university anyways. And even if they used a more popular language at university, chances are it's still not the one that would be actually useful for you.
|
yeah you guys were exactly right, awesome
and manitou: i think what you are saying may actually be a part of the reason they picked ocaml. They want something that's a little bit harder to get help with, lol. they are dicks like that. It seems this class is largely about reading documentation and learning new languages quickly (in 4 or 5 weeks we switch again, to prolog). However, the documentation for ocaml is abysmal to read.
|
On September 16 2017 21:25 travis wrote:
yeah you guys were exactly right, awesome
and manitou: i think what you are saying may actually be a part of the reason they picked ocaml. They want something that's a little bit harder to get help with, lol. they are dicks like that. It seems this class is largely about reading documentation and learning new languages quickly (in 4 or 5 weeks we switch again, to prolog). However, the documentation for ocaml is abysmal to read.
Wow. You can "learn" new language in a couple of days but it takes months to get semi-comfortable with it and years to actually understand it and know why some things are done this way rather than another. I really don't know what's the point of this course then. Learning new languages comes easier naturally with accumulated experience, not with force feeding them to people. If you only get 4-5 weeks of learning the language I can bet that in about 6 months you won't remember 99% of what you've learned (which, in my opinion, makes this entire exercise a waste of time). It's especially true for more esoteric languages (like OCaml and Prolog).
I've had a course in AI programming in Lisp at the university. It lasted entire year and I'm not sure I'd be able to write even a simple program in Lisp today without having to re-learn the basics. Good luck remembering something after just 4 weeks.
|
On September 17 2017 03:28 Manit0u wrote:Show nested quote +On September 16 2017 21:25 travis wrote:
yeah you guys were exactly right, awesome
and manitou: i think what you are saying may actually be a part of the reason they picked ocaml. They want something that's a little bit harder to get help with, lol. they are dicks like that. It seems this class is largely about reading documentation and learning new languages quickly (in 4 or 5 weeks we switch again, to prolog). However, the documentation for ocaml is abysmal to read. Wow. You can "learn" new language in a couple of days but it takes months to get semi-comfortable with it and years to actually understand it and know why some things are done this way rather than another. I really don't know what's the point of this course then. Learning new languages comes easier naturally with accumulated experience, not with force feeding them to people. If you only get 4-5 weeks of learning the language I can bet that in about 6 months you won't remember 99% of what you've learned (which, in my opinion, makes this entire exercise a waste of time). It's especially true for more esoteric languages (like OCaml and Prolog). I've had a course in AI programming in Lisp at the university. It lasted entire year and I'm not sure I'd be able to write even a simple program in Lisp today without having to re-learn the basics. Good luck remembering something after just 4 weeks.
I strongly disagree here.
At least not with the course being pointless. Yes, the language will neither be properly learned and even less will the language be remembered later on. But I don't think this is the goal anyway. (In fact, I don't believe teaching programming languages to a lasting effect is ever the goal at a university class...)
But given it's very special structure, really different from most classic languages, the course will still give insight in an entirely different approach to the whole topic of coding. Imho university classes are generally more focussed on putting you in different point of views, teaching your different approaches and general ideas how to tackle problems.
The real coding in the language of your choice is your job at home. This you have to do yourself. But learning how to properly think about the issue and how you could tackle the issue is the real key in those classes.
And I have seen enough people who know the languages, but don't know how to "think about coding". And the results of those are devastating. And working together in shared projects, and talking about how to do things is even worse.
And yes, today I don't remember a thing about my ultra short introduction into Haskell from almost a decade ago. But it still widened my horizon and this I actually kept. Same about PROLOG. I would never consider those classes a waste of time. I think it helped my skills, regardless the language I choose today.
|
On September 17 2017 04:33 mahrgell wrote:Show nested quote +On September 17 2017 03:28 Manit0u wrote:On September 16 2017 21:25 travis wrote:
yeah you guys were exactly right, awesome
and manitou: i think what you are saying may actually be a part of the reason they picked ocaml. They want something that's a little bit harder to get help with, lol. they are dicks like that. It seems this class is largely about reading documentation and learning new languages quickly (in 4 or 5 weeks we switch again, to prolog). However, the documentation for ocaml is abysmal to read. Wow. You can "learn" new language in a couple of days but it takes months to get semi-comfortable with it and years to actually understand it and know why some things are done this way rather than another. I really don't know what's the point of this course then. Learning new languages comes easier naturally with accumulated experience, not with force feeding them to people. If you only get 4-5 weeks of learning the language I can bet that in about 6 months you won't remember 99% of what you've learned (which, in my opinion, makes this entire exercise a waste of time). It's especially true for more esoteric languages (like OCaml and Prolog). I've had a course in AI programming in Lisp at the university. It lasted entire year and I'm not sure I'd be able to write even a simple program in Lisp today without having to re-learn the basics. Good luck remembering something after just 4 weeks. I strongly disagree here. At least not with the course being pointless. Yes, the language will neither be properly learned and even less will the language be remembered later on. But I don't think this is the goal anyway. (In fact, I don't believe teaching programming languages to a lasting effect is ever the goal at a university class...) But given it's very special structure, really different from most classic languages, the course will still give insight in an entirely different approach to the whole topic of coding. Imho university classes are generally more focussed on putting you in different point of views, teaching your different approaches and general ideas how to tackle problems. The real coding in the language of your choice is your job at home. This you have to do yourself. But learning how to properly think about the issue and how you could tackle the issue is the real key in those classes. And I have seen enough people who know the languages, but don't know how to "think about coding". And the results of those are devastating. And working together in shared projects, and talking about how to do things is even worse. And yes, today I don't remember a thing about my ultra short introduction into Haskell from almost a decade ago. But it still widened my horizon and this I actually kept. Same about PROLOG. I would never consider those classes a waste of time. I think it helped my skills, regardless the language I choose today.
I've never said that learning other languages is pointless. What is pointless is giving them just 4 weeks. It' not enough time in my opinion to really widen your horizons and teach you how to think about code. During this time you'll mostly just peruse the docs just to make the thing you're trying to do work. For the language to broaden your thinking you need to "get it" and understand what's so different about it from other languages you've been working with apart from the syntax. You can only do that once you're somewhat comfortable with the language and its structure, when you no longer need to focus on trivial stuff like "How do I properly iterate/recurse/map/reduce this structure here?".
|
On September 17 2017 05:09 Manit0u wrote:Show nested quote +On September 17 2017 04:33 mahrgell wrote:On September 17 2017 03:28 Manit0u wrote:On September 16 2017 21:25 travis wrote:
yeah you guys were exactly right, awesome
and manitou: i think what you are saying may actually be a part of the reason they picked ocaml. They want something that's a little bit harder to get help with, lol. they are dicks like that. It seems this class is largely about reading documentation and learning new languages quickly (in 4 or 5 weeks we switch again, to prolog). However, the documentation for ocaml is abysmal to read. Wow. You can "learn" new language in a couple of days but it takes months to get semi-comfortable with it and years to actually understand it and know why some things are done this way rather than another. I really don't know what's the point of this course then. Learning new languages comes easier naturally with accumulated experience, not with force feeding them to people. If you only get 4-5 weeks of learning the language I can bet that in about 6 months you won't remember 99% of what you've learned (which, in my opinion, makes this entire exercise a waste of time). It's especially true for more esoteric languages (like OCaml and Prolog). I've had a course in AI programming in Lisp at the university. It lasted entire year and I'm not sure I'd be able to write even a simple program in Lisp today without having to re-learn the basics. Good luck remembering something after just 4 weeks. I strongly disagree here. At least not with the course being pointless. Yes, the language will neither be properly learned and even less will the language be remembered later on. But I don't think this is the goal anyway. (In fact, I don't believe teaching programming languages to a lasting effect is ever the goal at a university class...) But given it's very special structure, really different from most classic languages, the course will still give insight in an entirely different approach to the whole topic of coding. Imho university classes are generally more focussed on putting you in different point of views, teaching your different approaches and general ideas how to tackle problems. The real coding in the language of your choice is your job at home. This you have to do yourself. But learning how to properly think about the issue and how you could tackle the issue is the real key in those classes. And I have seen enough people who know the languages, but don't know how to "think about coding". And the results of those are devastating. And working together in shared projects, and talking about how to do things is even worse. And yes, today I don't remember a thing about my ultra short introduction into Haskell from almost a decade ago. But it still widened my horizon and this I actually kept. Same about PROLOG. I would never consider those classes a waste of time. I think it helped my skills, regardless the language I choose today. I've never said that learning other languages is pointless. What is pointless is giving them just 4 weeks. It' not enough time in my opinion to really widen your horizons and teach you how to think about code. During this time you'll mostly just peruse the docs just to make the thing you're trying to do work. For the language to broaden your thinking you need to "get it" and understand what's so different about it from other languages you've been working with apart from the syntax. You can only do that once you're somewhat comfortable with the language and its structure, when you no longer need to focus on trivial stuff like "How do I properly iterate/recurse/map/reduce this structure here?".
Well... Imho the more different the language is, the less time is needed for the student to extract some new approach here. It isn't about mastery.
And given how much you have to bend your mind to "How do I properly iterate/recurse/map/reduce this structure here?" here, you are actually already learning here.
Most people, being thrown into a new language, will often produce code which looks almost like what they would have written in their language of origin. There are 2 big moments of learning, usually:
a) the moment you realize you can't do things the way you always have done, and now have to figure out how to do it here
b) the moment you realize, you can do things more efficient using the new language, instead of following your old ways.
b) is entirely out of reach for such super short introductions. And the moment you reach this point is probably the farther away, the more different the languages are.
But a) is the opposite. It actually happens much earlier, when the languages are super different. In this case, it already happens at the "How do I properly iterate/recurse/map/reduce this structure here?" stage. If the course only wants to grab this eye-opening moment.... This can be done in 4 weeks. Introduce it with some examples, let people repeat those simple examples, then let people apply those simple examples for slightly more advanced functions, like he posted above. Add an outlook, and you grabbed a good effect/time ratio.
And here it actually makes sense to pick some language which is more difficult to find c+p solutions for online. Because c+p is probably the killer of the entire a)-type learning. :D
|
On September 16 2017 01:11 Manit0u wrote:
I'd love to work on such project.
And regarding your classes: why the hell would they choose OCaml to teach you functional programming instead of something that's more widely used in the real world? There's Haskell, Erlang, Clojure... Technically, you could do functional programming in Ruby and Python too, you just wouldn't be using some of the language's features.
I think it's fairly debatable to say that Haskell is more widely used than OCaml in the real world, they're both pretty much niche.
Facebook both uses Haskell and OCaml btw. Haskell for their spam infrastructure iirc and OCaml for some static analysis tool.
|
On September 15 2017 22:45 travis wrote:
Anything crucial I should know? I know that everything is immutable in ocaml, which I guess is kind of the point with functional programming?
That's wrong. OCaml isn't pure.
You can write a counter function for example.
# let count = let x = ref 0 in let foo () = incr x; !x in foo;;
val count : unit -> int = <fun>
# count ();;
- : int = 1
# count ();;
- : int = 2
# count ();;
- : int = 3
# count ();;
- : int = 4
|
On September 16 2017 10:22 travis wrote:I think we use it because some of the school's sponsors push it anyways I already ran into a snag, and I am reading a lot about it but it's just so confusing to figure this out
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> true
This is fine. But not what I want. I want to see if h = t
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h::t -> (h = t)
not ok... "Error: This expression has type 'a list but an expression was expected of type 'a The type variable 'a occurs inside 'a list" ok so.. it evaluates to bool but I can't use it because it's not the same type as everything else? well then how the hell am I supposed to do this in a reasonable way?
Hanh already pointed it out, but tail is a list. I don't know the exact syntax of ocaml, but you'll want to do something of the kind:
let head_equal_tail lst = match lst with
[] -> false
| h::[] -> false
| h1::h2::t -> (h1 = h2)
That will return true if the first 2 elements of your list are equal (assuming the syntax is correct, which I cannot help you with), and otherwise false.
|
On September 17 2017 07:14 Liebig wrote:Show nested quote +On September 16 2017 01:11 Manit0u wrote:
I'd love to work on such project.
And regarding your classes: why the hell would they choose OCaml to teach you functional programming instead of something that's more widely used in the real world? There's Haskell, Erlang, Clojure... Technically, you could do functional programming in Ruby and Python too, you just wouldn't be using some of the language's features. I think it's fairly debatable to say that Haskell is more widely used than OCaml in the real world, they're both pretty much niche. Facebook both uses Haskell and OCaml btw. Haskell for their spam infrastructure iirc and OCaml for some static analysis tool.
i am a haskell programmer, there is a decent haskell community, ocaml seems popular in fintech though
|
On September 18 2017 06:56 aRyuujin wrote:Show nested quote +On September 17 2017 07:14 Liebig wrote:On September 16 2017 01:11 Manit0u wrote:
I'd love to work on such project.
And regarding your classes: why the hell would they choose OCaml to teach you functional programming instead of something that's more widely used in the real world? There's Haskell, Erlang, Clojure... Technically, you could do functional programming in Ruby and Python too, you just wouldn't be using some of the language's features. I think it's fairly debatable to say that Haskell is more widely used than OCaml in the real world, they're both pretty much niche. Facebook both uses Haskell and OCaml btw. Haskell for their spam infrastructure iirc and OCaml for some static analysis tool. i am a haskell programmer, there is a decent haskell community, ocaml seems popular in fintech though
Its always funny seeing OCaml randomly in the wild somewhere, for instance I just found out that one of the biggest unikernels MirageOS is written in OCaml.
Also do you use Haskell at your job?
|
Guys, I need a bit of help with conceptual work. Maybe some of you have had any experience with something like that before...
Basically it's all about process pipelining. We create the configuration and then each process step (step in the pipeline) receives some input files and produces some output files. The problem I'm facing at the moment is that different processes require different file and I need to map the correlations.
Example:
Process A has input files a, b, c and produces output files d, e.
Process B has input files f, g and produces output file h.
All the inputs/outputs need to be unique (it matters which one is which, don't ask me why) so when I go to the next step in the pipeline (process A finished) I need to know that I have to pass output d as input g and output e as input f so that process B can start.
I can't discover/calculate this on the fly since process configuration might change but all of the old jobs (series of steps) have to retain their old configuration.
Now. What do you think would be the best way to map such configurations in the database (postgres) so that it isn't too clunky to retrieve/parse?
The only thing I came up with this far is some form of config table, which will be attached to the process (one process has many configs and each config specifies the relation between output of this process and input of the next one) but this seems super fragile...
|
Canada16217 Posts
Anyone familiar with Moment JS? And knows how to parse dates/time pulled from an API(json format) into only hours/minutes?
|
On September 21 2017 08:56 Manit0u wrote:
Guys, I need a bit of help with conceptual work. Maybe some of you have had any experience with something like that before...
Basically it's all about process pipelining. We create the configuration and then each process step (step in the pipeline) receives some input files and produces some output files. The problem I'm facing at the moment is that different processes require different file and I need to map the correlations.
Example:
Process A has input files a, b, c and produces output files d, e.
Process B has input files f, g and produces output file h.
All the inputs/outputs need to be unique (it matters which one is which, don't ask me why) so when I go to the next step in the pipeline (process A finished) I need to know that I have to pass output d as input g and output e as input f so that process B can start.
I can't discover/calculate this on the fly since process configuration might change but all of the old jobs (series of steps) have to retain their old configuration.
Now. What do you think would be the best way to map such configurations in the database (postgres) so that it isn't too clunky to retrieve/parse?
The only thing I came up with this far is some form of config table, which will be attached to the process (one process has many configs and each config specifies the relation between output of this process and input of the next one) but this seems super fragile...
Why a postgres database? It sounds like you are trying to find a formal way of describing the structure of your relations, and not just the relations themselves. So you would want a language capable of representing and reasoning about the relational model of your database. That's a subset of first order logic, so mb look into that: by the sound of it, you could use prolog to do it fairly easily. You could also use OWL. These languages are built for reasoning about relations in a broader sense than SQL is. For instance, you can write the rules governing the system in the same language, for instance you can write in first order logic rules such as:
\forall b in Processes: \forall i in Wires: if input(, w) then (w in ProblemInputs or \exists b' in Processes \exists w' in Wires: b != b' and w != w' and output(b', w')
Or in language: all inputs of all your processes are properly connected. This can be done in Prolog and some flavors of OWL (which one is sometimes a bit tricky to figure out).
You can then verify that whatever configuration your user came up with satisfies the rules of the system.
|
|
|
|
|
|
EPT
