| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8
Comments:
<0> very free fall exams
<0> where you see something brand new and you need to apply stuff you learned
<0> but not in a trivial way
<0> many leap of faiths
<1> the most interesting sort of exams
<1> I both love and hate them
<0> exactly
<1> our high dimensional analysis lecturer basically told us the contents of our exams, which ****s
<0> the ones where you feel like a hero if you do it well
<1> last year, the analysis guy kept very quiet and set a hard exam. but I did well enough, so I was happy
<1> this year, it sounds like it'll be trivially easy, which is annoying. I know I'm good at the subject, and now I won't get a chance to prove it
<0> heh
<1> ah well. signed up for the "hard" analysis courses next year
<1> you must start your phd shortly? or are you waiting for sept to being?
<1> er, begin
<0> program analysis question that was kinda cool was to set up a flow algorithm that at each program statement tells you the minimum and maximum amount of times a definition of a variable will be used before being eliminated
<0> ie, live variables...
<1> sounds fun
<1> I look forward to examining the course in detail before I apply for it :)
<0> heh
<0> its easy to do the coursework etc for it
<0> all trivial
<0> but the exam
<0> is brain work
<1> indeed
<1> well, I'm sure you'll have done fine
<0> well see
<1> I'm expecting 6 easy exams and 2 hard exams at the end of the month
<1> suppose I should actually do some revision
<0> i know I didnt fail but that doesnt mean it went well
<0> I pulled an all nighter before my law exam
<1> I felt like that last year, and turned out to have done pretty well. people always seem to do better than they think
<0> ahem
<0> asriel, phd is in october
<1> fun summer planned then? :)
<0> but I got to finish my master dissertation now
<0> asriel, israel/london/us...
<1> israel, eh? sounds grand
<0> if iran doesnt start a meltdown... :)
<0> I need the beach
<0> havent had a good vacation in a looong time
<0> all I did in my uni life was internships over the summer
<0> but ill be in London over the summer too, my girlfriend will be there to finish her dissertation
<0> Ill be there to distract her
<1> :)
<0> her dissertation is about the common trend in the media to use 'stupid' blonde celebrities as idols, and how this impacts youth
<0> very touchy feely, not enough proofs
<1> heh :)
<1> I felt like that yesterday. doing a 2000 word essay on logic & proof systems
<1> other people are doing subjects like "The digital home" and other bs
<0> hehe
<0> yep
<0> tableau etc?
<0> did you also cover incompleteness theorem?
<0> I know its only 2000 words so very limited
<1> yup, incompleteness theorem is include
<1> d
<0> good man
<1> hilbert machines, propositional and predicate logic
<0> ah yes
<1> not written the tableau stuff yet, I'm not sure if they'll be room
<0> the good thing about imperial is that there is rarely a word limit on anything
<1> I was going for a sort of "Field introduction"
<1> there's technically not on this, but he said he wanted somewhere around 2000 words. I can probably push it to 2500
<1> the proper maths one next year has a limit ,but it's pretty large. somewhere north of 10,000 words
<1> plan was to mainly use propositional logic as a motivator. Can easily show, complete with reasonable riger, how propositional stuff can be used to prove something from first principles
<1> then introduce predicate stuff as an "upgrade" that fixes problems in the example
<1> then spend a few hundred words talking about existing implementations, and conclude
<0> http://www.imperialviolet.org/ that guy wrote a book with some profs in first year
<0> no second year
<1> heh
<1> reminds me, I need to get back to the people who gave me stuff to read
<0> heh
<0> pi calc stuff?
<1> aye
<1> well, it was their new book on ADPI
<1> just about finished (just reading about their method for distribution of processes)
<1> and now have enough knowledge to talk reasonablly about it, and reproduce some of the proofs
<1> most of which boil down to "We've got semantic axioms. Lets apply strutural induction!"
<1> so nothing very tricky mentally, just pages of tedious finickity bookeeping
<0> http://www.nytimes.com/2006/03/19/opinion/19crichton.html?ex=1300424400&en=9addb806498d2739&ei=5088&partner=rssnyt&emc=rss
<0> look at that
<0> crazy
<1> reading
<1> ah, yes. think I read this (or something very similar) when it was first published
<1> and yes, it's crazy. Hence why I spent quite a long time a year ago campaigning against the software/business methods EU patent directive
<1> http://www.archgrove.co.uk/weblog/category/politics/
<1> Though that doesn't include the 2 days of phone calls I made just before the vote. Didn't bother to write them up
<0> nice
<0> our law exam was about intellectual property
<0> in the EU the only way to get a patent on software at the moment is if it is embedded in a chip on a circuit
<1> in theory
<1> in practice, you can get them p***ed just by wording them abiguously
<0> our crypto exam was kinda fun, good trick question, imagine you use a hash function like this, take two blocks from the message, encrypt first one, xor it then with second one and encrypt the result again. why is this a very bad hash function
<1> not sure
<0> if you write down the RSA modular arithmetic you can see that you can very easily create another message with a clashing hash value
<0> hence its kinda worthless
<1> ah
<0> ugh :(
<1> on the plus side, they've introduced a new one on numerical linear algebra, which should be interesting
<1> but coding theory would have been more interesting
<1> ah well. c'est la vie
<0> indeed
<0> got full scholar ship for my phd, im a very happy man
<0> and my parents are happy too
<1> nice
<1> very nice :)
<1> bleh. within about 30 minute work of finishing this app, and my motivation is dead
<0> i need to start writing up the disertation and havent got the motivation either
<0> its all done in my head... writing up is boring
<2> is there any obvious reason why running a function via a .so would be much slower than running it directly as a binary?
<3> "in my end is my begining"(eliot)
<3> hi!!! telos
<4> telos: call through function pointer
<2> i'm calling the function from R by loading the .so .. that's the only way afaik
<5> Off to office
<6> #linuxhelp
<7> wtf
<8> hi, I have a array problem in c++, I have a cl*** member declared as double* m_ptFlow in the .h, in the .cpp I instantiate it doing m_ptFlow = new double[size]; then I use a for to load the data in the table this way .. table[i]=data. The problem is that the only data index in wich the table seems to keep a value for is m_ptTable[0], all the other indexes have 0 as value after the for, even if data is something else
<8> i'm realy not used to working with c++ and can't seem to figure the problem out
<1> sounds like your loop might be faulty
<1> post your code at the site in the topic
<8> it is a very simple loop..
<8> i'll post anyway
<7> for(int i=0; i<list; i++)
<8> ok i just posted it
<7> link?
<8> http://www.noidea128.org/sourcefiles/16258.html
<1> and after that for loop, your m_ptr has 0 in everything other than the first index?
<8> yes
<9> how do you know it? you print it?
<9> maybe the output logic is wrong
<1> looks fine to me
<1> so, as mathieu_ says, I suspect your debugging of it is the fault :)
<8> with a break point and a watch
<1> mm
<0> anyone seen garden state?
<8> so no ideas on my problem ?
<7> i looked at it.
<7> looks fine to me.
<1> logic looks fine. suspect the debugger might be deciving you
<1> or, the managed code compiler is mad
<8> I tried it on 2 different machines..
<9> make a second loop that prints the values... just to be sure you're debugging it correctly
<1> agreed
<9> because i can't see any error either
<9> (and ***ign 'i' to the elements.. because if it's the same loop with the same error, it will result in '2' each time anyway
<8> I I just tried no declaring double*m_ptTab as a cl*** member but as a local var for the constructor and it works fine..
<8> but when it is a cl*** member it dosen't work..
<1> could be a larger case of static/heap corruption
Return to
#c++
or
Go to some related
logs:
#chatzone
#chatzone
horneypot
wats op in london
#london
#linux
Linux GnuWorld Bot
#AllNiteCafe
#chatzone
#php
|
|