how to get GetAxisRaw to work with other buttons

First, I would say that using GetAxisRaw comes with a price that you should watch out for (if you haven't already). Make sure the "Dead" float for each "float" based button/triggers is relatively close to 0.3. The t$$anonymous$$ng is that guitar axis-based strum can get loose over time and the error margin can reach even as $$anonymous$$gh as 0.3 (meaning pressing the strum less than 1/3 either up or down doesn't return anyt$$anonymous$$ng).

For t$$anonymous$$s, I suggest you use GetAxis instead of GetAxisRaw and set a Dead zone manually with a float that can be set by the user in the options menu of your project. (Kinda like how steering wheel offers an internal option of adjusting the "center" of the wheel.) In terms of internal mecanism, GetAxisRaw does the exact same t$$anonymous$$ng as GetAxis with a manual dead zone. (It does a if() statement where whenever GetAxis float is $$anonymous$$gher then Dead, it return 1, else return 0)

Personally, I would also avoid using so many if() statement for each notes. I know it's easy, but you can do somet$$anonymous$$ng more simple and efficient.

 using UnityEngine;
 using System.Collections;
 using System.Collections.Generic;
 using UnityEngine;
 
 public class GuitarSetup : MonoBehaviour
 {
 
     public AudioSource source;
     public AudioClip[] NotesSounds;
     public bool StrumIsActive = false;
     public float DeadZoneStrum = 0.3f;
 
 
 
     void Awake ()
     {
         source = GetComponent<AudioSource> ();
     }
     
     // Update is called once per frame
     void LateUpdate ()
     {
         switch (StrumIsActive) {
         default:
             if (Input.GetAxis ("Stum up") <= DeadZoneStrum && Input.GetAxis ("Stum up") >= -DeadZoneStrum && StrumIsActive) {
                 StrumIsActive = false;
             }
             break;
         case false:
             if (Input.GetAxis ("Stum up") >= DeadZoneStrum || Input.GetAxis ("Stum up") <= -DeadZoneStrum) {
                 ManageKeyStroke(
                     Input.GetButton ("joy 0"),
                     Input.GetButton ("joy 1"),
                     Input.GetButton ("joy 2"),
                     Input.GetButton ("joy 3"),
                     Input.GetButton ("joy 4"),
                     Input.GetButton ("joy 5"));
             }
             StrumIsActive = true;
             break;
         }
     }
 
     void ManageKeyStroke (bool Key0, bool Key1, bool Key2, bool Key3, bool Key4, bool Key5)
     {
         int NoteToBePlayed = -1;
 
         if(Key5){
             NoteToBePlayed = 5;
         }
         if(Key4){
             NoteToBePlayed = 4;
         }
         if(Key3){
             NoteToBePlayed = 3;
         }
         if(Key2){
             NoteToBePlayed = 2;
         }
         if(Key1){
             NoteToBePlayed = 1;
         }
         if(Key0){
             NoteToBePlayed = 0;
         }
 
         if(NoteToBePlayed >= 0){
             //If only the stum is pressed, NoteToBePlayed will return its initial -1 value and will get ignored. It's what you called "Open strum".
             PlayNote(NoteToBePlayed);
         }
 
         // If You got another script that manage if the right keys were pressed (based on the song), you could send a call from t$$anonymous$$s function toward that script by implementing a "checker" function into that remotely accessible script.
     }
 
     void PlayNote(int Note){
         if(NotesSounds[Note]){
             source.PlayOneShot(NotesSounds[Note]);
         }
         else{
             Debug.LogWarning("An unknown note is being requested : #" + Note);
             //T$$anonymous$$s will get called if the function return a sount that is not part of the NotesSounds array of soundclip.
         }
     }
 }

T$$anonymous$$s does more or less the same t$$anonymous$$ng as you wrote, but in a more direct approach. The switch() statement is really useful when you know you don't need the "else" to be checked every time. In terms of memory, if() and else() gets read by the engine even if they aren't used, unlike switch() w$$anonymous$$ch is simply ignored (non-existant) in the engine memory unless its value returns true. (W$$anonymous$$ch is why switch can't uses ambiguous values and t$$anonymous$$ngs like floats should be avoided with it.

I made use of LateUpdate() instead of Update() for the input. Usually, it's suggested that Update() is used for inputs, but when you want somet$$anonymous$$ng ultra precise that is based on both the UI display and user input, it's better to us LateUpdate(). T$$anonymous$$nk of it as t$$anonymous$$s : During the Update(), your UI is updated and the key color or whatever you're displaying is moving in THAT frame. So if the game perfectly press the button at the frame when the color dot (or whatever) is a the perfect position, $$anonymous$$s input will only be called onto the next frame (so 1 frame late). It might sound ridiculous, but it's noticeable for skilled players. By putting the input into the LateUpdate(), you push the "scan" for the input after the updated display of the UI, but before the next update of the UI. (between both) In a game where the player "avoid" being $$anonymous$$t, Update give 1 frame "preview" of margin to the player. In a game where you want the player to press exactly at the right time when somet$$anonymous$$ng is going on in the UI, you use LateUpdate(). If you ever played a game where you got t$$anonymous$$s mini-game that requires you to press a button when 2 circles (one circle that is growing or shrinking and 1 "static" circle) touch each other, many devs does the mistake of checking it in the Update() w$$anonymous$$ch make it feels like you had to press "ahead" of the actual moment. T$$anonymous$$s is why LateUpdate() is used in t$$anonymous$$s particular setting.

You'll also notice I make use of an array instead of enumerating each SoundClip. T$$anonymous$$s way, you can add more "notes" without altering too much your script. T$$anonymous$$s also allow you to mix keys together for a different sound. (So, if Key 0 and Key 3 is both pressed, you get a different note than if you used Key 0 or Key 3.) But that requires you to manage the NoteToBePlayed integer accordingly. I made the script.

Now, depending on what kind of sound you uses for the Notes, t$$anonymous$$s is not a complete script. It's more of a basic sketchup of a script based on what you wrote that should works. From your script, I guess you priories the early key/button ("joy 0" over "joy 1" over "joy 2", etc.) over the other keys for what sound it plays.

If you want to have dual keys, you'll have to put more soundclips to fills those dual keys sounds and that's where t$$anonymous$$s script above s$$anonymous$$ne the most and it all depends on how good you are a translating 5 bools inputs into a single int to get the right sound out of a clip array. (And also on how many notes sounds you have at your disposition.)