Nyhetsflöde

A number of Master thesis opportunities are available through our guest lecturers. Please have a look at ir14 > Master thesis opportunities in the menu to the left.

Best,

Hedvig

Stämmer det alltså att föreläsningen 29 april är inställd?

Hi!

I just saw that the presentation was scheduled on May 18. However, when I looked at my calendar, May 18 is a sunday. Can I safely assume that this is a typo/something that have remain from last year? Because I really don't see the point of having a presentation on a weekend, where it has the greatest chance of interfering with people's plans (since most people probably didn't plan on this, seeing as it is not even an exam week). Especially not if presence is mandatory.

This was on the ir14 > Project page. Yes, it was a typo, it should be May 16!

This is also stated in ir14 > Schedule and course plan.

Ok, great, thanks! :)

I don't understand the logic behind this:

"Tip: If you are working on the CSC UNIX system, you do not have enough disc space on your account. You can then use the tmp/ directory on the local hard drive of the Linux machine you are using."

"Furthermore, the initial indexing step is very computationally expensive. It is natural that indexing of such a large data set takes several hours."

"NOTE: The indexing should have been done before the review session."

Where should we place our index when we do it before the review session? We cannot know which computer that will be assigned to us for the review session beforehand so we can't do the indexing like the day before. Even if we knew, /tmp might be cleaned the next day since it's temporary storage. I don't have enough space in my csc home directory. Am I required to bring my own laptop?

Ah! This is an issue. Back in a bit, it might be hard to find a solution before mon but I will try... Thanks. If you have a laptop, this is a good way to solve the problem.

Simple solution, publshed in news soon.

I only get 9 matches for the word "snowboard" for the first 10 000 documents, while according to the assignment, we should find 13 matches.

However, after running a "grep -i snowboard", I find 9 documents that has the word "snowboard" but 4 containing the word "snowboardåkare".

Didn't you verify that your Java environment tokenized the words correctly? I think it's ok you don't blame us if we have set up the environment incorrect, but you teachers should have made sure at least you have the correct settings before ;)

In fact, you should have written the code so that it reads files with the UTF-8 encoding (InputStreamReader with Charset UTF-8) instead of the platform default (FileReader). You should maybe also use decimal numbers instead of char literals for å ä ö etc. in the code to make sure the file is not compiled with the wrong charset. The same assignment (with some modifications) has been used the two previous years but you never corrected this? I have read many comments on KTH Social regarding people getting the "wrong" numbers and my friend tried to fix his "bugs" for five hours last Thursday while he actually had no bugs...

Snälla, titta på news, detta har vi gått igenom. Det står där OCH i lydelsen att några tal avviker. Det är inte till för att reta er. Dessa frågor är nya för i år och vi har jobbat otroligt hårt för att få allt klart i tid. Det är väldigt jobbigt och tråkigt att få sådana här arga kommentarer, så moderera dina inlägg i fortsättningen. Vi lärare är människor också...

Det var inte alls meningen att vara elak eller så utan snarare ett konkret tips på hur ni kan rätta till det här.

Jag förstår om ni har haft tidspress med att skapa labben, men oavsett så är det ganska irriterande när saker inte stämmer, vilket jag hoppas ni förstår...

Jag tror inte han menade illa. Jag kan hålla med några punkter Emil tar upp. För mig är det extremt viktigt att vara säker på att jag har gjort rätt, annars blir jag väldigt nojig inför redovisningen. Och det är svårt att veta om det står att man ska få andra siffror än vad jag får. Jag trodde att den här variationen av siffror gällde åäö, men kan ha missförstått det. Jag kan tänka mig att ni gör ert bästa att få det att bli så bra så möjligt. Det är bara just den biten som jag skulle uppskatta om det ses lite närmare på. :)

Ta inte illa upp, det är inte meningen att få er att må dåligt. :)

To answer the actual question: yes, Johan has rewritten the StringTokenizer since last year. It works on every OS - except Mac which I used to generate the answers. Since we were in a hurry, we found out only when the students started. For assignment 2 we are generating answers on Linux.

There is usually a natural explanation to things, so do not start by assuming that people are stupid/lazy.

Aha, I thought the same tokenizer was used this year as last years, and therefore you had observed different results depending on environment the other years as well (and therefore should have fixed the problem for this year). But since you have a new tokenizer this year that was not the case :)

Dear Emil och Simon,

Emil, thanks.

Of course I do not think it is ok with errors in the assignment, and we vere very angry with ourselves when this was discovered.

The reason why I was upset now was 1) that exactly the example above had been discussed at length a couple of days ago, so Emil just had to scroll to get this info; 2) the tone of voice in the post. A tip: avoid irony smileys in texts to people you do not know, they have the opposite effect than intended. But happy smileys are ok! :)