imputeformants: Impute Missing Formant Values

imputeformantsR Documentation

Impute Missing Formant Values


Impute missing formant values using a least-squares approximation.


imputeformants (ffs, speaker, vowel)



A numeric vector containing formant frequency measurements for a single formant. Values should be in Hz, and missing values need to be set to 0.


A vector indicating which speaker produced each formant.


A vector indicating which vowel category each formant belongs to.


This function finds the least-squares approximation for each missing value based on the assumption that each formant for a given vowel differs between-speakers solely on the basis on a speaker-specific multiplicative parameter. This assumption is well supported in the literature (Nearey 1978, Nearey & Assmann 2007, Turner et al. 2009). This parameter would be most closely related to gross speaker vocal-tract length.


A vector containing each original formant frequency and imputed formant frequencies where appropriate.


Santiago Barreda <>


Nearey, T. M. (1978). Phonetic Feature Systems for Vowels. PhD thesis, Indiana University Linguistics Club.

Nearey, T. M. & P. F. Assmann. (2007). Pobabilistic 'sliding template' models for indirect vowel normalization. in Experimental Approaches to Phonology, edited by M.J. Sole, P. S., Beddor, and M. Ohala (Oxford University Press, Oxford), pp. 246-269.

Turner, R. E., Walters, T. C., Monaghan, J. J. M., & Patterson, R. D. (2009). A statistical, formant-pattern model for segregating vowel type and vocal-tract length in developmental formant data. The Journal of the Acoustical Society of America, 125(4), 2374. doi:10.1121/1.3079772


#data (h95)

## Select F2 values from the tenth speaker in H95 data
## set the first 5 values to "missing"
#ffs = h95$f2
#ffs[h95$speaker == 10][1:5] = 0
#speaker = h95$speaker
#vowel = h95$vowel

## impute these missing values
#imputedf2 = imputeformants (ffs, speaker, vowel)

## resulting in a very close approximation of the original values
#plot (imputedf2[h95$speaker == 10], h95$f2[h95$speaker == 10])
#abline (0, 1, col = 2)

phonTools documentation built on Nov. 21, 2023, 1:07 a.m.