Hello, this is what I found on an old french forums topics elsewhere on the web.

Don’t know if it is correct though:

// John F. Carter Squeeze

//////////////////////////////////////////////////////////////////
// Période des Moyennes Mobiles (défaut = 20)
nPer = 20

// Coefficient des Canaux de Keltner (défaut = 1.5)
fKeltner = 1.5

// Coefficient d'écartement des Bandes de Bollinger (défaut = 2)
fBB = 2.0

// Canaux de Keltner
aKeltnerUp = Average[nPer](TypicalPrice) + (fKeltner * AverageTrueRange[nPer])
aKeltnerDw = Average[nPer](TypicalPrice) - (fKeltner * AverageTrueRange[nPer])

// Bandes de Bollinger
aBollingerUp = Average[nPer](Close) + (fBB * STD[nPer](Close))
aBollingerDw = Average[nPer](Close) - (fBB * STD[nPer](Close))

// L'indicateur vaut 1 sauf si les Bandes de Bollinger sont
// comprises entre Canaux de Keltner
nSqueeze = 1
If (aBollingerUp <= aKeltnerUp) and (aBollingerDw >= aKeltnerDw) Then
nSqueeze = -1
EndIf

// Delta of price from Donchian mid line
DonchianDelta = LinearRegression[nPer] ( Close - ( (Highest[nPer](High) + Lowest[nPer](Low)) / 2 + ExponentialAverage[nPer](Close))/2 )

// Conseil : afficher DonchianDelta sous forme d'histogramme et
// nSqueeze sous forme de Points d'épaisseur maximum
Return DonchianDelta As "DonchianDelta", nSqueeze As "CarterSqueeze"

I remember having coded something similar myself, but I don’t remember for who and when, I think it was a private job for one of the customer of the programming services, so I can’t share it.

For information, I coded the TTM Squeeze indicator for prorealtime, this is the link for it into the library.

Sorry but why do you think it is not the correct one? What are the noticeable differences please?

I can remember a color inversion some talked about, let me check please.