As with most languages, reading Python's error messages is a skill that needs to be practiced. Also like with other languages, it's a skill worth learning.
The error messages should be read like this:
- The last line states what type of error occurred (
TypeError in this case) and gives a (hopefully) descriptive message. In your case, there were too many arguments given to the mfccInitFilterBanks function.
- Jumping up to where the error message starts (
Traceback...), we read down to see the chain of what's being done from highest level (the statement in your code) to lowest level (the function's definition in the package)
- We look for where
mfccInitFilterBanks was called, which is the lowest level: in the package code of pyAudioAnalysis, line 669 of the audioFeatureExtraction.py file (here's the GitHub version).
In that package file, mfccInitFilterBanks is defined like this:
def mfccInitFilterBanks(fs, nfft):
"""
Computes the triangular filterbank for MFCC computation (used in the
stFeatureExtraction function before the stMFCC function call)
This function is taken from the scikits.talkbox library (MIT Licence):
https://pypi.python.org/pypi/scikits.talkbox
"""
However, that same file later calls the function like this:
[fbank, freqs] = mfccInitFilterBanks(Fs, nfft, lowfreq, linsc, logsc, nlinfil, nlogfil)
So the problem's with the pyAudioAnalysis package. You should create an issue on its GitHub page.