The following are some mails I've sent you.

---------email #1--------

Dear all,

The hw description is at the end of the .R file which I displayed in class. You can found it on Prof. Kuan's website or my google drive (recommended)(https://docs.google.com/folder/d/0B1CkHS88RyBMUEJ1QjBqaUtWZG8/edit)

Hw1 is due on 10/6 23:59:59, please upload it to ceiba before deadline. You can hand it in late, but this will reflect on your score. Please demonstrate what you can do as much as possible.

Feel free to ask.

Ying Chin

---------email #2--------

Dear all,

Some of you ask me some questions about programming homework, I'll explain it more clearer:

You have to choose something you want to test for and state it clearly in the beginning of your .R file. For instance, if you'd like to know if beta_1 + beta_2 = 0 or not, write the null hypothesis in the beginning. The next and most important thing is to consider how to design your own function? What is the R and r in equation (3.16) in lecture notes? (Though programming is usually not show up in midterm, understanding this will help you A LOT!)

Some of you email me, saying that calling the function: summary(lm( . )) , then R will show "F-statistics: ......" at the last line of its summary. I want to remind you that the underlying null hypothesis of most of the statistical packages is "H0: beta_1+beta_2+...+beta_k=0", which may not be the statement that you want to test. So google a built-in R function that can help you test exactly what you're interested in, and compare with the function you code by yourself.

NOTE:

Do the homework on your own! The correctness is not the only grading criterion, so put yourself to the furthest!

Good luck:D:D:D

Ying Chin


---------email #3--------

Dear all,

I found that there are two ceiba websites for the two different departments (and cannot combine them into one), which make it annoying for sending notices for all of you guys. Thus let's forget about poor ceiba! I'll email you according to this list and put the content of my mails into a .txt file named "Announcement" in my google drive, just check it any time. While for convenience, you should still upload your homework to ceiba and I'll grade there.

sleepy, 
Ying Chin


------------mail#4--------------

Dear all,

I have to correct the previous hint I told you: The null hypothesis of f-test in most of the statistics packages is "H0: b1=b2=...=bk=0", which is actually "H0: b1=0, and b2=0, and ... ,and bk=0". That is the R in equation 3.16 is a k*k identity matrix.

Stupid,
Ying Chin O_Q