or
:99binomcdf(127,0
:augment(Ans,augment(Ans,Ans->L1
:0

Edit: Right number of augments now.

can someone explainthis in detail; so i can use id

it seems to be very use full and how do you go about compressing it??

I was playing around with some of the other variables.
A matrix (32*3) is exactly 875 bytes.
Now, is there a way to store complex numbers to said matrix?

Found a pretty good compression trick.
(Remember, I'm trying to do this all in TI-BASIC. Don't want all my games to require external APPS.)

If you store your list as a string to one of the sequence variables, u, v, w, you can half the size of the entire thing.
Then, if you need to evaluate it, run 'u' and Ans(X where X is the list element.

The only issue with this is the lack of editability if that's a word.

I have a list of exactly 384 elements.
Said list is ~3900 bytes.
Compressed into complex numbers is ~1950 bytes.

What kind of elements are you saving (integers, decimals)?  The most efficient type of compression to use will rely on that.

In general, to get the most out of list compression it helps to understand how the calculator stores certain things:
1) Real numbers are always exactly 9† bytes, and store exactly 14† digits.
2) Complex numbers are nothing but two real numbers put together; they're 18† bytes and can store two 14-digit numbers.
3) Lists are just a bunch of real/complex numbers stored sequentially.  A four-element real list will be 4*9=36† bytes; a two-element complex list (basically four numbers) will be 2*18=36† bytes.
3b) There is no such thing as a list with real and complex elements; if a list has one complex element, then every other element is treated as such and takes up 18 bytes, even if they have no imaginary part.
4) Matrices are basically lists.  An NxM matrix is just an N*M element list of real numbers, so a 4x5 matrix is 4*5*9=180† bytes.
4b) The OS doesn't support complex values in matrices, only real numbers.  The reason for this is probably that matrices can quickly grow to over 10KB without the user realizing it, so doubling the size to allow for complex numbers would kind of suck.

Considering (3) and (3b), how did you use complex numbers to halve the size of the list?  By adding complex list elements, you'll halve the number of elements, but the remaining ones will double in size.

†This is for the data itself.  For an actual variable (e.g. L1 or A), there are a few extra bytes containing the variable name and some other stuff.

---

To get said 384 element list:

Code: [Select]

99binomcdf(127,0->L1
augment(L1,L1
augment(Ans,L1->L1
0


If you want a 384 element list of all 99's, you could just use (assuming you want L1):
:384→dim(L1
:Fill(99,L1This also has the nice effect of not storing the result in Ans, so you don't need double the spare RAM (however, storing the dimensions of the list does modify Ans).

---

Matrices can be multiplied by a (real) number, but you can't add number to one.

I remember multiplying by a number worked. I don't think adding by a number did though.

Oh, you are right.

To add a number to a matrix, you need to create another matrix of the same size, fill it with the number you want to add/subtract, and then you can add/subtract those matrixes.  I use this in Shift, ans so i had a dedicated addition matrix

TI-BASIC turns quite fun with all these works around. Still a nice technique.

I haven't found in real games any usefulness of row+( and row*( commands... It would be curious to see one.

If you store your list as a string to one of the sequence variables, u, v, w, you can half the size of the entire thing.
Then, if you need to evaluate it, run 'u' and Ans(X where X is the list element.

The only issue with this is the lack of editability if that's a word.

That's neat.

I don't use complex numbers for compression because I don't see anyway to make things smaller with them.