Hey all,

I have been trawling the net and scratching my head for a couple of hours now:

I have a text file with about 120 words in it. I want to randomise the word order within pd and output that as another text file.

What would be the best way of going about this?

Many thanx in advance,

Jb

Well I sure wish there was a tutorial on this sort of stuff, but the only help I can point you to is located in Help > Browser > 5.reverence > list-abs/ > *

But if/when you find the solution, PLEASE do post your learned knowledge here. Thank you.

Cheers Arif,

Atm I'm thinkin about using 'random' 'select 1-120' and writing each word in a message box going into print. Which seems well clunky:(

If I can find something a little more elegent I will defo post.
Bests,

Jbz

Okay, my first glance Idea, involves turning a text file into a list (somehow), you might need to [list-enumerate] so that you can then [list-accum] and get the total number of entries. Create an array with maxsize being the total number of entries and resolution being enough to count every word, then you will later [shuffle] between 0 and total number of entries, to get your random index value at every drop of the textfile into [drip] ((starting to get lost, myself)) then, you generate a random number for word number and place that as the value into your table [tabwrite] going along 1 by one, (using random index value) then you have a shuffled table. and you just need to work out some way to play back that information as words

Right then.

I've used 'list-enumerate' and I now have a text file that has each word assigned a number with a semicolon on the end of each word too. I'm trying to put that into an array but each of the items in the array list has a value of 0. I'm not sure what you mean by 'resolution' of the array. Could you explain further, also, what is 'list-accum' being used for.

Ta for this,

Jb

how about storing the words in [coll] or [textfile] objects, then calling them randomly? It'd be a lot easier than connecting "120" cables! ...

Or [msgfile] from zexy accepts an index number

tcihan's idea of [coll] sounds good to me. Attaching the output of [list-enumerate] to [coll] should fill it up, and you can then just send a random number to [coll] and it will output the word associated with the number.

You might also try [list-idx], which takes an index number in its right inlet and a list in the left, then outputs the element associated with the index number (no need to enumerate). I think [coll] will be faster, though.

Many thanx to you guys who weighed in,

So, I think I'm almost there.

I used list-idx and now have my word list being spat out in a random order. Joy.

Now I could just cut and paste the output of print, but I would like to do this 'properly' and be able to create a new text file with just the wordlist, without having 'symbol' as a prefix.

Any pointers peeps?

Jbz

[add $1(
|
[textfile]

Hey all,

Ok, I attach my version for others to expand and improve.

Big props to those that helped.

I'm gonna do it by hand for now but it would be really good to have the various words taking up a whole line rather than being in a list where there is one word per line, also, of course, with a space between each word. Getting pedantic now though:)

Cheers y'all,

Jbz

http://www.pdpatchrepo.info/hurleur/word_manipulation.pd

[add2 $1( for one line.

After taking a quick glance at your patch, I think it might smooth things out to have a [textfile] at the top of the patch also and read your text file into that. I think as long as you don't have any semi-colons in the text file, it should spit every word out as a list (not entirely sure, though). Then you can bang it out into [list-len] to get the number of words into [shuffle]'s right inlet. Then send the list into the right inlet of

to store it instead of copying and pasting into that message box.