Movatterモバイル変換

[0]ホーム

version 0.5.3

NEWS.md setup

Version 0.5.4 2025-07-14

‘diarisation’ substituted by ‘identification’ in DESCRIPTION.

Version 0.5.3 2025-07-14

Drop https://cloud.r-project.org/bin/macosx/ from vignette because‘CRAN URL not in canonical form’.
Drop https://hf.co/settings/tokens fromdiarize.Rbecause it is ‘Unauthorized’ by CRAN.
Drop https://www.ee.columbia.edu/~dpwe/resources/matlab/rastamat/fromextract_features.R because it is ‘Forbidden’ byCRAN.
Drop https://apps.dtic.mil/sti/pdfs/ADA303368.pdf fromextract_features.R because it is ‘Forbidden’ by CRAN.

Version 0.5.2 2025-07-11

Title case: Voice Analysis, Speaker Recognition and Mood Inferencevia Music Theory.
Added functions sourced from theembedr package byMichael McCarthy, under MIT License:https://github.com/mccarthy-m-g/embedr/blob/master/LICENSE.md. Thisinclusion is temporary and will be discontinued onceembedris available on CRAN. See https://github.com/mccarthy-m-g/embedr formore details.
Omitted+ file LICENSE from DESCRIPTION (CRAN).
Shifted ‘Speaker Recognition’ and ‘Voice Analysis’, so The Titlefield DO NOT starts with the package name (CRAN).

Version 0.5.1 2025-07-10

Added to DESCRIPTION file: Imports: arrangements, Suggests: embedr,gm, tidyverse.
Removed from DESCRIPTION file: Imports:gm.
Functionspn2octave renamed tospn2abc,examples updated (http added,octave removed).
TODO done: update octave > spn2abc @ check_chords.R.
Added#' @param x @ assign_notes.R.
Added.data$ before f0, f1 and f2 @ piano_plot.R.
Added\dontrun to check_chords.R examples whentry_perm = TRUE. Take a long time!

Version 0.5.0 2025-07-10

Correcting minor issues for publication.

Version 0.4.31 2025-04-11

Correcting errors.
Minor errors corrected.

Version 0.4.30 2025-04-11

Functionoctave.R renamed tospn2octave.R.
Added functioncheck_chords.R.

Version 0.4.29 2025-03-16

Added functionoctave.R.

Version 0.4.28 2025-03-14

Imports `ggplot2``.
Addsggplot2:: ataes,xlab,ylab andggtitle.

Version 0.4.27 2025-03-14

Minor corrections.

Version 0.4.26 2025-03-14

Package title updated to ‘Voice Analysis, Speaker Recognition andMood Inference via music theory’.
Removed'mfcc','rf','rcf','rpf','rfc'from the default argument ofextract_features.R
Added'gain' to the default argument ofextract_features.R
Added functionpiano_plot.R.
Added functioncut_audio.R.
Added functionassign_notes.R.

Version 0.4.25 2024-06-09

Addedtry to line 209 ofsplitw.R.

Version 0.4.24 2024-03-25

'pyannote/speaker-diarization' set to'pyannote/speaker-diarization-3.1' @ diarize.R
Addedspeechbrain tohttps://raw.githubusercontent.com/filipezabala/voice/master/requirements.txt

Version 0.4.22 2022-MM-DD

pycall set from'~/miniconda3/envs/pyvoice38/bin/python3.8' to'~/miniconda3/envs/pyvoice/bin/python' @ diarize.R andextract_features.R

Version 0.4.21 2023-04-20

Dropped the section 3 fromvoicegnette_R after ProfBrian Ripley’s email: “Dear maintainer, Please see the problems shown onhttps://cran.r-project.org/web/checks/check_results_voice.html.Please correct before 2023-05-03 to safely retain your package on CRAN.The CRAN Team”

Version 0.4.20 2023-04-20

filesRange andfeatures switched. NowfilesRange is in the 3rd andfeatures in the2nd positions in the argument vector. This change impacts the examplesofinterp_df.R,interp_mc.R,smooth_df.R,tag.R.
f0_praat, …,f8_praat implemented inextract_features.R.
pycall = '~/miniconda3/envs/pyvoice38/bin/python3.8'added as argument inextract_features.R.
formants set tofmt inextract_features.R.
windowShift set to10 as default inextract_features.R andfeat_summary.R.
extract_features_py.R deprecated.
voice::conv set tovoice::interp at lines45 and 50 ofinterp_mc.R.
Functionsget_ removed fromspltw.R.
More information given invoice::diarizedocumentation.

Version 0.4.19 2023-04-06

All documentation verified.
voice::conv set tovoice::interp.
voice::conv_mc set tovoice::interp_mc.
voice::conv_df set tovoice::interp_df.
gender set tosex as argument ofvoice::extract_features.
gender set tosex as argument ofvoice::feat_summary.
gender set tosex as argument ofvoice::tag.

Version 0.4.18 2023-04-05

Addedpycall argument tovoice::extract_features_py.
References updated invoice::extract_features.
References updated invoice::extract_features_py.
F0 column inferred viavoice::extract_features_py set tof0_praat.Line 39 of extract_f0.py, (df_f0_long['f0_praat']) and line83 of extract_features_py.R(dplyr::select(id, file_name, interval, F0) %>%).
1, column inferred viavoice::extract_features_py set tof0_praat.
Columnmhs set tof0_mhs invoice::extract_features.
Lean call tovoice::extract_features:features = c('f0', 'formants', # Pitch and formants'df', 'pf', # Formant dispersion and position'rf', 'rpf', 'rcf', # Formant removals'rfc', # (R)e(F)lection (C)oefficients'mfcc'), # (M)el (Frequency (C)epstral (C)oefficients
RPf - Formant Position Removal by Zabala (2023) disentangled from Pfby Puts et al (2012) invoice::extract_features.
voice::diarize calls directlypyannote.audio viareticulate using a token.The download and call oflibs.py anddiarization-pyannote.py is no longer needed.
Reconnect to Github after 5 months.

Version 0.4.17 2023-03-14

voice now depends of R (>= 4.0.0), not R (>= 4.1.0) in orderto allow Kaggle installation.
Deprecatedpoetry (sent todraft/poetry.R), createdR/diarizeinstead.
Deprecateddata/id_path.rda andman/id_path.Rd (sent todraft/id_path.rda anddraft/id_path.Rd), created/data/mozilla_id_path.rda andman/mozilla_id_path.Rd instead.
Updated .Rbuildignore: ^.*.Rproj$ ^.Rproj.user$ ^cran-comments.md$^CRAN-SUBMISSION$ ^requirements.txt$ ^draft$ ^revdep$
rstudio.com became posit.co.
Monterey became Ventura.
voice::diarize exported to NAMESPACE.

Version 0.4.16 2022-09-14

Allow R/conv_df.R to use unitaryfeaturesargument.

Version 0.4.15 2022-09-07

Removed duplicated ‘must’ fromautoDir parameterdocumentation at poetry.R
Updated vignettes/voicegnette_R.Rmd.
Setverbose = FALSE:
- R/feat_summary.R
- R/tag.R
Setto.data.frame = to.data.frame tocn.licall @ R/conv_df.R.

Version 0.4.14 2022-09-02

As suggested by Benjamin Altmann:
- Added@return to R/write_list.R. This implies adding a\value field to the corresponding .Rd file.

Version 0.4.12

As suggested by Benjamin Altmann:
- Reduced the length of the title to 64 characters, less than 65. InDESCRIPTION:
  nchar('Tools for Voice Analysis, Speaker Recognition and Mood Inference')
- Package names, software names and API names written in single quotesin title and description.
- enrich_rttm.R: ‘’
- extract_features_py.R: ‘Python’s’ ‘Parselmouth’
- feat_summary.R: ‘voice::extract_features’
- poetry.R: ‘Python’s’ ‘pyannote-audio’
- DESCRIPTION: ‘R’ and ‘Python’
Reference “Zabala, F.J. (2022) to appear in…” removed from
- extract_features.R
- feat_summary.R
- tag.R
Added@return to the following .R files regardingexported methods and explaining the functions results in thedocumentation. This implies adding a\value field to thecorresponding .Rd files.
- R/audio_time.R
- R/enrich_rttm.R
- R/extract_features.R
- R/extract_features_py.R
- R/feat_summary.R
- R/get_bit.R
- R/get_dur.R
- R/get_left.R
- R/get_right.R
- R/get_samp.rate.R
- R/get_tbeg.R
- R/get_tdur.R
- R/is_mono.R
- R/notes.R
- R/poetry.R
- R/read_rttm.R
- R/splitw.R
- R/tag.R
Functions moved to draft directory:
- draft/chords.R
- draft/has_audio.R
- draft/id_file.R
- draft/is_can.R
- draft/is_ext.R
- draft/memory.R
- draft/na_filter.R
- draft/notes_summary.R
- draft/plot_note.R
- draft/plot_q.R
- draft/rowProp.R
- draft/rp.R
- draft/spoken_time.R
Functions in which\dontrun was replaced with or simplyadded\donttest:
- R/conv_df.R
- R/conv_mc.R
- R/enrich_rttm.R
In order to easily suppress information messages to the console,argument ‘verbose’ was set to as default @ R/extract_features.R .
tempdir() used @:
- R/enrich_rttm.R
- R/get_tbeg.R
- R/get_tdur.R
- R/poetry.R
- R/read_rttm.R
- R/splitw.R
\url added to references at
- R/enrich_rttm.R
- R/notes_freq.R
- R/notes.R
- R/read_rttm.R

Version 0.4.11 2022-08-28

In order to reduce the examples running time:
- The following variables were omitted from conv_df.R example:‘zcr’,‘mhs’,‘rms’,‘gain’,‘rfc’,‘ac’,‘cep’,‘dft’,‘css’,‘lps’,‘mfcc’.
- \dontrun added to conv_mc.R.
- \dontrun added to na_filter.R.

Version 0.4.10 2022-08-28

“You also have to remove the licences file which is part of Ranyway. Best, Uwe Ligges”. License file removed byunlink('LICENSE') command.
License set to the original ‘GPL-3’ (instead of ‘MIT + fileLICENSE’, ‘MIT’ or ‘GPL-3 + file LICENSE’) @ DESCRIPTION file.

Version 0.4.9 2022-08-26

License set to ‘MIT + file LICENSE’ (instead of either ‘MIT’,‘GPL-3’ or ‘GPL-3 + file LICENSE’) @ DESCRIPTION file.

Version 0.4.8 2022-08-26

License set to ‘MIT’ (instead of either ‘GPL-3’ or ‘GPL-3 + fileLICENSE’) @ DESCRIPTION file.

Version 0.4.7 2022-08-26

License set to ‘GPL-3’ (instead of ‘GPL-3 + file LICENSE’) @DESCRIPTION file. As I remove it, a new note was presented:
‘File LICENSE is not mentioned in the DESCRIPTION file.’

Version 0.4.6 2022-08-25

Now depends: R (>= 4.1.0) in order to install on Windows.

Version 0.4.5 2022-08-24

Following Uwe Ligges’ list.
URL: https://www.apple.com/macos/big-sur/ moved tohttps://www.apple.com/macos/monterey/ @ README.md.
Using fully specified URLs starting with the protocol forhttps://www.xquartz.org/ @ README.md.
Changed http –> https for https://ffmpeg.org/ andhttps://www.fon.hum.uva.nl/praat/ @ README.md. Notehttp://filipezabala.com/ and http://www.rob-mcculloch.org/ arenon-secure protocols.
The canonical URL of the CRAN page was updated forhttps://cran.r-project.org/package=wrassp,https://cran.r-project.org/package=tuneR,https://cran.r-project.org/package=seewave andhttps://cran.r-project.org/package=gm.
Can’t find where to omit “+ file LICENSE”, the DESCRIPTION filerequires it.

Version 0.4.4 2022-08-23

Thelist.txt file was generated bywrite_list.R function.\dontrun environmentwas applied to solve the problem atdevtools::check.

Version 0.4.3 2022-07-18

Replaced \url by at 10.1121/1.421048 and 10.1098/rspb.2011.0829 @extract_features.R.
Where is list.txt?????

Version 0.4.2 2022-07-18

Removed “≈” U+2248 Almost Equal
Adjusted Maintainer and Authors@R (Zabala Filipe J.): Maintainer:‘Zabala Filipe J.filipezabala@gmail.com’Authors@R: ‘Zabala Filipe J.filipezabala@gmail.com’
Where is list.txt?????

Version 0.4.1 2022-07-18

Many improvements after devtools::release().
Where is list.txt?????

Version 0.4.0 2022-07-17

Many improvements after devtools::check().
Where is list.txt?????

Version 0.3.22 2022-07-15

Update documentation @ extract_features.R.

Version 0.3.21 2022-06-10

Lost changes.

Version 0.3.20 2022-05-30

Corrected verbose = TRUE @ tag.R and @ feat_summary.R.

Version 0.3.19 2022-05-30

Added @ voice::extract_features in line 22: #’@param verbose Logical.Should the running status be showed?
Verbose also added to feat_summary.R and tag.R.
feat_summary2 set to feat_summary in line 31 @ feat_summary.R.

Version 0.3.18 2022-05-23

Added dplyr::tibble in line 92 @ feat_summary.R.

Version 0.3.17 2022-05-22

Solved summarizing problem when wav_path was a directory (and not afile name) @ feat_summary.R.

Version 0.3.16 2022-05-22

Lost changes.

Version 0.3.15 2022-05-21

Updated feat_summary.R normalizing dirnames @ Media and @ Extended(full path)
Added extdata directory.

Version 0.3.14 2022-05-07

Lost changes.

Version 0.3.13 2022-MM-DD

Set id_path.Rda to mozilla_id_path.Rda

Version 0.3.12 2022-02-13

RemovedtoLowerGroupBy @ feat_summary.R and tag.R.

Version 0.3.11 2022-02-11

Set check.mono and stereo2mono to FALSE @ extract_features.R,feat_summary.R and tag.R.
Addedx <- dplyr::as_tibble(x) @feat_summary.R.

Version 0.3.10 2022-02-11

Set wavFiles <- do.call(rbind, wavFiles) to wavFiles <-do.call(rbind, as.list(unlist(wavFiles))) @ extract_features.R.

Version 0.3.9 2022-02-10

Tidy up tag.R documentation, removing@param subj.id,@param media.id and@param subj.id.simplify.
Added@paramwavPath in feat_summary.R and tag.R documentation.
Added@paramwavPathName in feat_summary.R and tag.R documentation.
Normalized dirnames @ feat_summary.R

Version 0.3.8 2022-02-09

dplyr::vars @ feat_summary.R

Version 0.3.7 2022-02-09

Make Media and Extended datasets consistent in documentation.
extract_features.R
- Drop ‘file_name_ext’ column @ output.
- Add ‘wavPath’ as the third column @ extract_features.R output.
- Admit directories or files in the main argument x.
feat_summary.R
- path > wav_path, id_seq > slice_seq, id_seq_file >slice_seq_file
- Added ‘groupBy’ argument
- Check tag.R applied on Canonical data
tag.R
- Create examples
- Remove arguments subj.id, media.id and subj.id.simplify
- Disable audio_time

Version 0.3.6 2022-01-02

Added@references Ardila et al (2019).
Added voice::plot_q.R, a plot_q[uality] function.
Imports magrittr at DESCRIPTION file.

Version 0.3.5 2022-01-02

Removed voice_ex1.rda, 7.5MB.
Deleted NAMESPACE.

Version 0.3.4 2022-01-01

Code:id_path <- E %>%select(client_id:path)save(id_path, file = '~/Dropbox/D_Filipe_Zabala/pacotes/voice/data/id_path.rda', compress = 'xz')voice::id_path

Version 0.3.3 2022-01-01

Code:Eid <- E %>%select(client_id:path)save(Eid, file = '~/Dropbox/D_Filipe_Zabala/pacotes/voice/data/id_path.rda', compress = TRUE)data(voice::id_path)

Version 0.3.2 2021-12-15

Harmonize voice::extract_features.R and voice::tag.R arguments.

Version 0.3.1 2021-12-03

extract_features2 set to voice::extract_features @ line 24 @feat_summary.R.

Version 0.3.0 2021-12-03

Set argument ‘directory’ to ‘x’ @ extract_features.R.
Set features colnames to lowercase @ extract_features.R.
Added df, pf, rf, rcf and rpf as argument of ‘features’ @extract_features.R.
Deprecate argument ‘extraFeatures’ @ extract_features.R and,feat_summary.R and tag.R.
ContemplateZCR,RFC,AC, andMFCC (numbers) @ feat_summary.R.

Version 0.2.11 2021-11-27

Added dplyr::bind_cols to line 602 @ extract_features.R.

Version 0.2.10 2021-11-27

RCf <- paste0('RCf', 2:8)
RPf <- paste0('RPf', 2:8)

Version 0.2.9 2021-11-27

Formant Dispersion Removal (RDf) set to Formant Cumulated Removal(RCf), line 53 @ feat_summary.R
dat$Dfn <- dat$Fn-dat$F1 set todat$Dfn <- (dat$Fn-dat$F1)/(n-1) @ Df - FormantDispersion by Fitch (1997) @ extract_features.

Version 0.2.8 2021-11-27

Formant Dispersion Removal (RDf) set to Formant Cumulated Removal(RCf).
RCf set from i=1,…,8 to i=2,…,8.

Version 0.2.7 2021-11-10

Added suppressMessages to line 61 @ smooth_df.R.
Summary generalized to all features obtained fromvoice::extract_features @ voice::feat_summary.

Version 0.2.6 2021-11-07

Added'rule = 2' to line 52 @ conv.R.

Version 0.2.5 2021-10-25

Added voice::audio_time to voice::tag.
New tag order @ voice::feat_summary: mean, sd, vc, median, iqr,mad.

Version 0.2.4 2021-10-14

Addedtag_F0_iqr = IQR(F0, na.rm = TRUE) @feat_summary.R.
Addedtag_F0_mad = mad(F0, na.rm = TRUE) @feat_summary.R.

Version 0.2.3 2021-10-13

Setx[,subj.id] <- as.character(cumsum(!duplicated(x[,subj.id])))@ tag.R.

Version 0.2.2 2021-10-13

Added parameter subj.id.simplify and all voice::extract_featuresparameters @ tag.R.

Version 0.2.1 2021-10-13

tag.R incorporates parametermc.cores = 1 and dropped i= 4.
set tag.R from feat_summary2 to feat_summary at line 26.
feat_summary.R running voice::extract_features instead ofextract_features.
feat_summary.R return(M_summ) instead of return(M).

Version 0.2.0 2021-10-13

Added has_audio.R.
Added audio_time.R.
Added spoken_time.R
Added feat_summary.R
Added tag.R.

Version 0.1.1 2021-09-27

audio_id.R set to id_file.R.
audio_id(x, i = 5, drop_fn = FALSE) set toid_file(x, col = NULL, pattern = '[_.]', i = 5, drop_col = FALSE)
id set toid_seq,id_file settoid_seq_file @ extract_features.R.

Version 0.1.0 2021-09-26

extract_features creates a general id and an id_file, counting byfile.
‘formant nested removal’ (RNf) set to ‘formant dispersion removal’(RDf) by Zabala (2021/2022) @ extract_features.R and @ data.R.
Removed person(“Salum Jr.”, “Giovanni A.”, email =“gsalumjr@gmail.com”, role = “aut”).
splitw must show the elapsed time
drop_fn = TRUE set todrop_fn = FALSE

Version 0.0.0.9047 2021-09-21

audio_id.R function created.
Exported voice_ex1.rda and data.R.
toFile = F set to toFile = FALSE @ extract_features.R.
features.list[[i]] <- dplyr::tibble() set tolapply(features.list, dplyr::tibble) @ extract_features.R.
Added ‘formant dispersion’ (Df) by Fitch (1997)(doi:10.1121/1.421048)
Added ‘formant position’ (Pf) by Puts, Apicella & Cárdenas(2011) (https://doi.org/10.1098/rspb.2011.0829)
Added ‘formant removal’ (Rf) by Zabala (2021/2022) @extract_features.R.
Added ‘formant nested removal’ (RNf) by Zabala (2021/2022) @extract_features.R.
Added ‘formant position removal’ (RPf) by Zabala (2021/2022) @extract_features.R.

Version 0.0.0.9046 2021-09-15

Added.name_repair = 'unique' totibble::as_tibble @ smooth_df.R. This solved the Warning:Thex argument ofas_tibble.matrix() must haveunique column names if.name_repair is omitted as of tibble2.0.0. Using compatibility.name_repair.

Version 0.0.0.9045 2021-09-15

names(rttm) <- basename(rttmFiles), or file/basenames inserted as names in list @ read_rttm.R.
Added columns~black and~Black tonotes_freq.R.
Addedmethod == 'black' andmethod == 'Black' @ notes.R.
Added distance <- diff(freq) @ notes.R.
method set tomeasure @ notes.R.
frequency set tofreq @ notes_freq.R. .New col order @ notes_freq.R: freq, spn.lo, spn.hi, spn, midi, black,Black, wavelength .as_tibble set totibble::as_tibble @ smooth_df.R. .xs_li settosnum_li @ smooth_df.R. .xs_df set tosnum_df @ smooth_df.R. .xs_final set toxs @ smooth_df.R. . Ordered Imports: dplyr, e1071, ellipse,ggfortify, ggplot2, R.utils, RColorBrewer, reticulate, seewave, tibble,tidyr, tuneR, VIM, wrassp, zoo

Version 0.0.0.9044 2021-09-12

Parameters fromWav, fromRttm admits either file or directory @splitw.R.

Version 0.0.0.9043 2021-09-12

Added parameteras.tibble = TRUE @ enrich_rttm.R.

Version 0.0.0.9042 2021-09-12

Changed column name from ‘id.split’ to ’id_split @enrich_rttm.R.
is.num <- unlist(lapply(x, class)) == 'numeric'changed tois.num <- sapply(x, class) %in% c('integer', 'numeric')@ conv_df.
Added smooth_df.R.
Drop by.filter parameter @ conv_df.R.
id and colnum shifted @ conv_df.R documentation.
Added parameteras.tibble = TRUE @ enrich_rttm.R.

Version 0.0.0.9041 2021-09-10

Changedmc.cores = parallel::detectCores() tomc.cores = 1 @ conv_df.R, conv_mc.R,extract_features.R.
Argumentsilence.gap set from 0.4 to 0.5 (universalstandard) @ splitw.R, enrich_rttm.R.
Changed column name from ‘id.min.time’ to ‘id.split’ @enrich_rttm.R.

Version 0.0.0.17

Solved the recursive issue in diarization-pyannote.py (bySola):

import osimport re

file_list = []for root, dirs, files in os.walk('./'):file_list = [f for f in files if re.findall(r'[.]wav', f, re.IGNORECASE)]break

print(file_list)

os.listdir?
os.walk(pathfrom, level=1)?
os.walk(pathfrom, recursive = FALSE)?
https://stackoverflow.com/questions/229186/os-walk-without-digging-into-directories-below

Version 0.0.0.1 2018-11-21 08:30:00 GMT-3

[8]ページ先頭