Data processing

Some functions are offered to help summarize the data generated when you analyze a folder. At the moment there are just some basics. Output goes into a directory called 'processed_data' in your working folder.

Aggregate

This function will aggregate the data in your 'csvs' folder into a single csv file. There will be one row for each sound, indicating its duration, and its mean/median f0. In addition, formant frequencies are aggregated at equally-spaced time points, using either the median or the mean. The number of time points is determined by the number of bins. For example, selecting 5 bins divides formant track into 5 equal sections, and finds the median/mean inside each section.

Get coefficients

Information about regression coefficients for each formant for each sound is stored in the info files (in the 'infos' folder). This function collects all these and puts them into a single CSV file. These coefficients are useful for all sorts of reasons, including being useful in modelling average formant trajectories across all speakers and in the automatic classification of speakers and speech sounds.