diamond.square.terrain

[jenessalemon]

@Wolflab
I am having a hard time with writing a function to go through the matrix applying the diamond and square step sequentially. Here is what I want to do in English:

1. Apply the diamond and square step to the whole matrix (which I've done.)
2. Somehow index each quadrant individually and apply the diamond and square step to those.
3. Then I think I just want to keep getting quadrants of quadrants until the matrix is full. (Which means that I have to have a certain kind of matrix, not JUST an odd one, right?)

So step two is what I really need help with. I think that I know the basics of indexing a matrix (matrix[row, column]) but I don't know how to index a whole block of a matrix. @willpearse?

[willpearse]

Looking good! You're getting better at writing things out first. Once you've written something out as you do above, try getting to the next level and writing something that looks a bit more like pseudo-code. A good way of telling if you've written pseudo-code is whether or not you use words like for or sapply in what you're writing. You're coming along leaps and bounds, though - well done!

Make sure you're applying the diamond step then the square step to all sub-matrices/quadrants.

Consider whether step 1 and step 2 are really different steps. In one you're applying to a particular sub-matrix (which just happens to be the whole matrix!) and in the second you're looping over a load of sub-matrices.

As far as indexing goes, you're already thinking about the right lines, so let me give you a hint. You are correct in that you know how to get out a single row and a single column (--> a single cell in your matrix). See if you can figure out how to do it using the example of how to do something similar in a vector below:

vector <- c("Will", "Sarah", "Jeni", "Barry")
toupper(vector[2])
toupper(vector[2:3])

...what do you reckon?...

[jenessalemon]

In order to apply diamond before square in every sub-matrix, can we just use a for loop?
Ex: for (i in crazy index)
sapply diamond step
sapply square step
I thought that R will execute the commands in order?

As far as indexing goes, I do know how to get chunks of a vector like that. It is, however, harder to work in 2D! I am not very good at programming so I decided to try a for loop. I'm sure there is a better way.
I wrote this little function to see if I could get the first quadrant and it was a fail.
quad <- function(matrix){
n <- length(matrix)
for (i in 1:median(n)){
4 <- matrix[i : median(n)]
}
}
quad(setup)

This is the error message:
Error in 4 <- matrix[i:median(n)] :
invalid (do_set) left-hand side to assignment
@willpearse I've never seen that before, can you help?

[willpearse]

Look at what I've written (vector[2:3]), look at what you've written (matrix[row,col]) and see if you can find a way to join both of those things up. I cover this kinds of subsetting in the first lecture.

The thing with errors is to read them slowly and carefully. I'll help you break down this one.

1. "Error in 4 <- matrix[i:median(n)] : " --> the line 4 <- matrix[i:median(n)] has an error in it. Read the line out loud to yourself. What do you think it does?
2. "invalid (do_set) left-hand side to assignment". You probably don't know what the bit in parentheses means, but it's in parentheses so it mustn't be that important (else it wouldn't be in parentheses - that's just grammar :D). So the key phrase is "invalid left-hand side to assignment"
3. Breaking that sentence down, there's clearly something wrong with the left hand side of your code. "assignment" probably refers to something being given a value, which sounds like the <- operator to me. So it's what's to the left of that.

I would imagine that's enough to get it...

[Wolflab]

Will: You beat me to it by seconds. I was about to say that Jenessa can fix this one, especially your item 1) the error message.

[jenessalemon]

Can you index the index? Like as in:
matrix[row[1:median(n)], col[1:median(n)]] ?
If so I might have an idea. I figured out that the error message was because I was trying to assign 4 a value, when I needed to be assigning a cell to be 4. Sorry I am so slow at this.

[willpearse]

I wouldn't call it indexing the index, but yes you can. Go back to your
handout and look at the first session on matrices.
On Sat, 12 Nov 2016 at 23:20, jenessalemon [email protected] wrote:

Can you index the index? Like as in:
matrix[row[1:median(n)], col[1:median(n)]] ?
If so I might have an idea. I figured out that the error message was
because I was trying to assign 4 a value, when I needed to be assigning a
cell to be 4. Sorry I am so slow at this.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#4 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/ABLcUoxOPo3-dEKyb456ahDFRQtpa5kyks5q9kmpgaJpZM4Kwhyl
.

[jenessalemon]

Ok, so matrix[1:3, 1:3] grabs the first quadrant! Now it's a matter of trying to apply both steps to all four quadrants, and all four of their sub quadrants, and so on, which I believe I am going to use expand.grid to use.

[jenessalemon]

It's my turn, I'm trying to call you on Skype