Based on the pandas dataframe data structure, CAMAT (Computer-aided Music Analysis Tool) offers convenient possibilities both for the comparative statistical analysis of several sheet music files and for searching for musical units and patterns. In the context of the Fellowship project, these possibilities could be implemented only exemplarily so far and are limited in the current version to the comparison of pitch classes and intervals as well as to the search for interval sequences. They are explained in the tutorials using music examples.

* Tutorial CAMAT_SheetMusic_Advanced_Part3_Corpus_statistics (Jupyter notebook; html-preview)

The tutorial introduces possibilities for queries of several note files, e.g. the different movements of a multi-movement composition.

The comparative queries focus on the following characteristics of each voice or all voices of one or several sheet music files:

• range in measures and time signatures used
• ambitus incl. lowest and highest note
• frequencies of the 12 pitch classes (absolute value and percentage)
• Frequencies of the interval steps (up and down, absolute value and percentage)

The results of the queries are displayed in tables in the browser (cf. html-preview), where you can choose between absolute frequencies and relative frequencies (percentages) for the frequencies.

* Tutorial CAMAT_Basics_Part4_Search (Jupyter notebook; html-preview)

This tutorial is dedicated to searching for interval sequences (independent of rhythmic design) in the individual voices of a sheet music file.

The input of the interval sequences to be searched is has to be entered as semitone steps, e.g. [2, 2, -4] (= two second steps up followed by a major third down).

The output is given as a list of the occurring instances with indication of the starting note, the voice, the measure and the position in the measure. In addition, the frequencies of the results on different starting tones (pitch classes) can be displayed.