I read some posts saying it's not possible to convert a symbol to a float.
However I found out there is a way and it's not even complicated.

It took me hours though to try all the other ways that don't work.
So for anyone who is trying the same, here's how to do it:

1. put your symbol into [symbol2list]
(optionally using a known delimiting character)
2. use [unpack f]
(to select the first (or only) element)
or
use [unpack ? ? f ? ?]
(to select a special list element you want to extract)
3. continue to work with the f output of unpack which is a float now
(if you chose the right delimiting character and element for your input of course)

hints and tips are always welcome. cheers.

[l2s] aka [list2symbol] is also good if u want to go the other way, and convert a float into a symbol.

hi
i followed your instructions like that:

[symbol abc(
I
[symbol2list]
I
[unpack f]
I
[print]

and it gave me following:
error: unpack: type mismatch

it only worked with unpack s

what did i do wrong??
greets
flx

i think this is only supposed to work with symbols that have number inside them

i confirm this works ! i thought it was impossible and would allways avoid the pb !

you can also use [fromsymbol]

I had a problem with converting a list of separated floats to a single float (4 9 2 to 492). Another issue I had was that the list/digit size was variable. The solution I used was : l2s (symbol 0 in right inlet) >> s2l >> unpack [f].

Thank you for this,

been arsing around trying to find a way and bingo. Yay.
Funny thing is I realise that I never seem to use symbol boxes at all.
Got some OSC messages that need parsing so here we are:)

Props,

Julian

I have been trying to implement this but when i unpack f it splits my number into the individual numbers i.e 1475 comes out as 1 4 7 5 into separate number boxes. I know rem-d mentioned putting symbol and the appropriate delimiter which I have tried but still not succeeded. Does anybody know of any ways to combine 4 floats back into a new number?

Hi Sophie, try this:

Cheers!
Gilberto

@gsagostinho Took me too long to post!!
here it is (more or less the same as Gilberto) in a patch..........join_floats.pd
David.

@whale-av hehe you need to patch faster, David!

(BTW, soon I will finish editing a video of my work with reacTIVision and then I will send you the link. In the end it worked really well, but had to be indoors in order to control the light)

@gsagostinho Woof.. no, I need slower competition!
I am glad your work went well in the end... looking forward to the link!

Very useful indeed, thanks for posting this!

Hello,

I know that this is very old thread but I also reached here yesterday so I think this might be useful to someone like me!

I am using MobMuPlat(=MMP) + Pd and MMP can get user string input and Pd also can receive it.
If I type sth. like -34.123456 then Pd also can get it but in string not as a number!

To convert this back to a proper floating point number, I have managed to implement a working object for my situation.

Hope this help someone else in the future.

tofloat.pd

It is using [list symbol] to get series of numbers first, and detect characters one by one to do needed actions with several variables ([expr]), and then finally spits out integral part & fractional part.

Happy patching!

BTW, this is my first post to the Pd community.
I really love Pure Data.
Thank you for making it & supporting it !

Chao,

Dooho Yi

Hi @applecargo
Thanks for sharing your achievement through your first post in this forum.
For some releases now, Pd can convert symbols of numeric values to float by simply passing the symbol through a [float] object. For examples:

[3.14\ <- this is a symbol (Ctl+4)
|
[f ]
|
[3.14\ <- this is a number atom (Ctl+3)

This won't work with symbols in messages though, from what I've tested.

@alexandros @applecargo [f] will work to convert from Pd 0.46 with messages but the symbol tag must be explicitly added as it is normally stripped by a message box......
As ever...... precision always lost when passing the message.
David.

@alexandros Hi, thank you for the tip! Oh, this was that easy- O-o;;

@whale-av I won't investigate more on my case, but I strongly feel like this was my case. It wasn't prepended by 'symbol'. Good to know. This would be useful later. Thanks.

@alexandros said:

[3.14\ <- this is a symbol (Ctl+4)
|
[f ]
|
[3.14\ <- this is a number atom (Ctl+3)

Since 0.5? shortcut ctrl+4 has been replaced by the new list-box. Symbol-box has no shortcut anymore, it's still in the 'put-menu'.

@lacuna That is correct, still, such a patch will work the same way. The [float] object can convert symbols to floats from the list-box too. Still, a good point.