rhythm.metrics is an R package for calculating and
visualising speech rhythm metrics from consonantal and vocalic interval
durations.
The package provides functions for computing widely used rhythm measures, including Delta C, Delta V, Varco C, Varco V, %V, rPVI-C, and nPVI-V, as well as plotting functions for exploring and presenting results.
Note:
rhythm.metricsis under active development, and additional metrics may be added in future releases.
You can install the development version of
rhythm.metrics from GitHub:
install.packages("remotes")
remotes::install_github("congzhang365/rhythm.metrics")library(rhythm.metrics)The package expects a data frame containing interval labels, utterance identifiers, and interval durations.
A typical input data frame looks like this:
df <- data.frame(
cv_label = c(
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel"
),
utterance_id = c(
"utt_1", "utt_1", "utt_1", "utt_1",
"utt_2", "utt_2", "utt_2", "utt_2",
"utt_3", "utt_3", "utt_3", "utt_3",
"utt_4", "utt_4", "utt_4", "utt_4"
),
cv_duration = c(
0.10, 0.80, 0.20, 0.50,
0.30, 0.30, 0.40, 0.70,
0.30, 0.88, 0.50, 0.90,
0.30, 0.57, 0.40, 0.97
),
utterance_duration = c(
2.4, 2.4, 2.4, 2.4,
2.7, 2.7, 2.7, 2.7,
3.4, 3.4, 3.4, 3.4,
1.8, 1.8, 1.8, 1.8
)
)| Category | Function | Description |
|---|---|---|
| Calculation | delta_cv() |
Calculate Delta C and Delta V |
| Calculation | varco_cv() |
Calculate Varco C and Varco V |
| Calculation | percentage_v() |
Calculate percentage of vocalic intervals (%V) |
| Calculation | rpvi_c() |
Calculate raw Pairwise Variability Index for consonants |
| Calculation | npvi_v() |
Calculate normalised Pairwise Variability Index for vowels |
| Plotting | plot_delta_cv() |
Plot Delta C and Delta V |
| Plotting | plot_varco_cv() |
Plot Varco C and Varco V |
| Plotting | plot_percentage_v() |
Plot %V |
| Plotting | plot_rpvi() |
Plot rPVI-C |
| Plotting | plot_npvi() |
Plot nPVI-V |
Delta C and Delta V are rhythm metrics based on:
Ramus, F., Nespor, M., & Mehler, J. (1999). Correlates of linguistic rhythm in the speech signal. Cognition, 73(3), 265-292.
delta_cv(df, cv_label, utterance_id, cv_duration)plot_delta_cv(df, cv_label, utterance_id, cv_duration)Varco C and Varco V are based on:
Dellwo, V. (2006). Rhythm and Speech Rate: A Variation Coefficient for deltaC. In P. Karnowski & I. Szigeti (Eds.), Language and language-processing (pp. 231-241). Peter Lang.
varco_cv(df, cv_label, utterance_id, cv_duration)plot_varco_cv(df, cv_label, utterance_id, cv_duration)%V is based on:
Ramus, F., Nespor, M., & Mehler, J. (1999). Correlates of linguistic rhythm in the speech signal. Cognition, 73(3), 265-292.
It measures the percentage of total utterance duration occupied by vocalic material.
percentage_v(df, v_label = "vowel", utterance_id, cv_duration, utterance_duration)plot_percentage_v(df, cv_label, label_name = "vowel",
utterance_id, cv_duration, utterance_duration)rPVI-C is based on:
Grabe, E., & Low, E. L. (2002). Durational variability in speech and the rhythm class hypothesis. In Laboratory Phonology 7 (pp. 515-546). De Gruyter Mouton.
It calculates the average absolute difference between consecutive consonantal intervals.
rpvi_c(df, cv_label, label_name = "consonant", utterance_id, cv_duration)plot_rpvi(df, cv_label, label_name = "consonant", utterance_id, cv_duration)nPVI-V is based on:
Grabe, E., & Low, E. L. (2002). Durational variability in speech and the rhythm class hypothesis. In Laboratory Phonology 7 (pp. 515-546). De Gruyter Mouton.
It calculates the normalised average absolute difference between consecutive vocalic intervals.
npvi_v(df, cv_label, label_name = "vowel", utterance_id, cv_duration)plot_npvi(df, cv_label, label_name = "vowel", utterance_id, cv_duration)A simple workflow with the package might look like this:
library(rhythm.metrics)
df <- data.frame(
cv_label = c(
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel",
"consonant", "vowel", "consonant", "vowel"
),
utterance_id = c(
"utt_1", "utt_1", "utt_1", "utt_1",
"utt_2", "utt_2", "utt_2", "utt_2",
"utt_3", "utt_3", "utt_3", "utt_3",
"utt_4", "utt_4", "utt_4", "utt_4"
),
cv_duration = c(
0.10, 0.80, 0.20, 0.50,
0.30, 0.30, 0.40, 0.70,
0.30, 0.88, 0.50, 0.90,
0.30, 0.57, 0.40, 0.97
),
utterance_duration = c(
2.4, 2.4, 2.4, 2.4,
2.7, 2.7, 2.7, 2.7,
3.4, 3.4, 3.4, 3.4,
1.8, 1.8, 1.8, 1.8
)
)
# Analysis
delta_cv(df, cv_label, utterance_id, cv_duration)
varco_cv(df, cv_label, utterance_id, cv_duration)
percentage_v(df, v_label = "vowel", utterance_id, cv_duration, utterance_duration)
rpvi_c(df, cv_label, label_name = "consonant", utterance_id, cv_duration)
npvi_v(df, cv_label, label_name = "vowel", utterance_id, cv_duration)
# Visualisation
plot_delta_cv(df, cv_label, utterance_id, cv_duration)
plot_varco_cv(df, cv_label, utterance_id, cv_duration)
plot_percentage_v(df, cv_label, label_name = "vowel",
utterance_id, cv_duration, utterance_duration)
plot_rpvi(df, cv_label, label_name = "consonant", utterance_id, cv_duration)
plot_npvi(df, cv_label, label_name = "vowel", utterance_id, cv_duration)For more detailed examples and usage notes, see the package vignette and function help pages after installation.
?delta_cv
?varco_cv
?percentage_v
?rpvi_c
?npvi_vIf you use rhythm.metrics in your research, please
cite:
Zhang, C. (2022). A Guide for the R Package “rhythm_metrics”. OSF Preprints. https://doi.org/10.31219/osf.io/kfnzt
You can also obtain the package citation from R with:
citation("rhythm.metrics")BibTeX for the 2022 guide:
@misc{zhang2022,
title = {A Guide for the R Package "rhythm_metrics"},
author = {Zhang, Cong},
year = {2022},
doi = {10.31219/osf.io/kfnzt},
url = {https://doi.org/10.31219/osf.io/kfnzt}
}Bug reports, feature requests, and suggestions are welcome. If you encounter an issue or would like to suggest an additional metric, please open an issue on GitHub or email me at cong.zhang@newcastle.ac.uk
GPL-3