README.md

In fauxneticien/kphon: Helper functions for the Kaytetye Phonological Corpus

kphon

Helper functions for the Kaytetye Phonological Database (kphon).

Installation

# If you don't have the devtools package, then do 'install.packages("devtools")' first
devtools::install_github("fauxneticien/kphon")

Examples

Create standardised export CSV and wav files for kphon, which are of the form:

| orthography | gl_annotated | transcription | annotator | kphon_export | annotation_source | xmin | xmax | | :------------- | :------------- | :-------------- | :-------- | :------------------------------- | :-------------------------------- | -------: | -------: | | atnyemilparele | south.wind-ERG | ɐcɲejmejlpɐɻələ | MP | atnyemilparele_AR_cf125a3a.wav | southwindpassage_AR_MP.TextGrid | 2.125691 | 3.140994 | | kape | and | kɐpə | MP | kape_AR_29cf0b23.wav | southwindpassage_AR_MP.TextGrid | 3.223741 | 3.606705 | | aherrkele | sun-ERG | ɐːɾkələ | MP | aherrkele_AR_4bcff7be.wav | southwindpassage_AR_MP.TextGrid | 3.606705 | 4.075386 | | tyampe | too | cɐmpə | MP | tyampe_AR_95ac9af5.wav | southwindpassage_AR_MP.TextGrid | 4.075386 | 4.511355 | | ahenge | fight | ɐːŋə | MP | ahenge_AR_e67432f2.wav | southwindpassage_AR_MP.TextGrid | 4.660752 | 4.950565 |

Step 0: Load kphon library and set working directory

library(kphon)

setwd("~/sandboxes/kphon-v2/annotations/southwind/")

Step 1: Provide TextGrid file(s) to process, with an Input/Output mapping

# Step 1: Provide input TextGrid file(s)
#         and an input-output 'io' mapping for TextGrid tiers to columns
step1 <-
    textgrid_to_kphon(
        input_files = "southwindpassage_AR_MP.TextGrid",
        io_mapping = list(
            orthography   = "word",
            gl_annotated  = "gloss",
            transcription = "phon"
        )
    )

Step 2: Derive (however appropriate) speaker and annotator columns, then feed into hash function to derive file name for export

step2 <-
    step1 %>%
    mutate(
        speaker      = str_extract(string = annotation_source, pattern = "_[A-Z]+_") %>% str_remove_all("_"),
        annotator    = str_extract(string = annotation_source, pattern = "[A-Z]+\\.TextGrid") %>% str_remove("\\.TextGrid"),
        input_wav    = str_replace(string = annotation_source, pattern = "_[A-Z]+.\\TextGrid", ".wav"),
        output_wav   = hash_kphon_wav(orthography, speaker, annotation_source, xmin)
    )

Step 3: After you’ve verified data extracted correctly in Step 1 and Step 2, write .wav and .csv files

write_kphon_exports(kphon_df = step2, path = ".", overwrite = TRUE)

fauxneticien/kphon documentation built on Aug. 1, 2019, 3:09 a.m.