def plot_from_range_dict(range_dict,
pdf_path,
nr_fft=100,
len_fft=1024,
nr_mel_bins=100):
"""
Plot the ranges in the range_dict in mel spectrum scale and save to file
Parameters:
-----------
range_dict : dictionary
Dictionary whose keys are the filenames and values are list of lists of start and end time
for the training interval ranges.
pdf_path : string
Path to a location where a pdf file is created containing the plots for all
ranges in range_dict
nr_fft : int
The number of ffts in each stft calculation
len_fft : int
The length of each of the nr_fft fourier transforms for each stft
nr_mel_bins : int
The number of bins in which the mel spectrum is to be divided in
Results:
--------
None
"""
X = intvpck.train_interval_mel_features(range_dict,
nr_fft=nr_fft,
len_fft=len_fft,
nr_mel_bins=nr_mel_bins)
plot_titles_lst = pckhlp.plot_titles_list_from_dict(range_dict)
pckhlp.save_feat_in_pdf(X, (nr_mel_bins, nr_fft), pdf_path,
plot_titles_lst)
return None
def plot_from_range_dict(range_dict,pdf_path,nr_fft = 100,
len_fft = 1024, nr_mel_bins = 100):
"""
Plot the ranges in the range_dict in mel spectrum scale and save to file
Parameters:
-----------
range_dict : dictionary
Dictionary whose keys are the filenames and values are list of lists of start and end time
for the training interval ranges.
pdf_path : string
Path to a location where a pdf file is created containing the plots for all
ranges in range_dict
nr_fft : int
The number of ffts in each stft calculation
len_fft : int
The length of each of the nr_fft fourier transforms for each stft
nr_mel_bins : int
The number of bins in which the mel spectrum is to be divided in
Results:
--------
None
"""
X = intvpck.train_interval_mel_features(range_dict, nr_fft = nr_fft,
len_fft = len_fft, nr_mel_bins = nr_mel_bins)
plot_titles_lst = pckhlp.plot_titles_list_from_dict(range_dict)
pckhlp.save_feat_in_pdf(X, (nr_mel_bins,nr_fft), pdf_path,plot_titles_lst)
return None
def compute_range_dict(source_file_path, pattern, config_par, fs):
"""
Compute the range dictionary for a single audio file for a given pattern.
Paramters:
----------
source_file_path : string
Path to the input wav file with sample frequency fs.
pattern : ndarray
The timeseries pattern which is used to calculate the correlation with data. In sample frequency fs.
config_par : dict
Dictionary containing all the required configuration parameter
fs : float
The sample frequency (frames per second) of the data
Returns:
--------
range_dict : dictionary
Dictionary whose keys are the filenames and values are list of lists of start and end time
for the training interval ranges.
roll_max_peaks : ndarray
Rolling maximum of data normalised by its rolling mean.
"""
# -- Calculate the average correlation over all freq_range_fft_bins between source and pattern
peaks = correlation_picking(source_file_path,
pattern,
config_par['freq_range_fft_bins'],
len_fft=config_par["len_fft"],
nr_ffts_per_s=config_par["nr_ffts_per_s"],
pattern_len_s=config_par["pattern_len_sec"])
# -- Identify frames of isolated peaks
centre_frames, roll_max_peaks = peak_identification(
peaks,
width_in_s=config_par["peak_mute_width"],
width_roll_mean=config_par["width_roll_mean"],
roll_max_peaks_threshold=config_par['peaks_threshold'],
fs=fs,
nr_ffts_per_s=config_par["nr_ffts_per_s"],
chunk_len_s=60,
len_fft=config_par['len_fft'],
is_ret_roll_max_peaks=True)
print 'centre_frames', centre_frames
# -- Create range dictionary around isolated peaks
range_dict = intvpck.training_range_times_from_peaktimes(
centre_frames, config_par["range_len_sec"], source_file_path, fs)
return range_dict, roll_max_peaks
def compute_range_dict(source_file_path, pattern, config_par ,fs):
"""
Compute the range dictionary for a single audio file for a given pattern.
Paramters:
----------
source_file_path : string
Path to the input wav file with sample frequency fs.
pattern : ndarray
The timeseries pattern which is used to calculate the correlation with data. In sample frequency fs.
config_par : dict
Dictionary containing all the required configuration parameter
fs : float
The sample frequency (frames per second) of the data
Returns:
--------
range_dict : dictionary
Dictionary whose keys are the filenames and values are list of lists of start and end time
for the training interval ranges.
roll_max_peaks : ndarray
Rolling maximum of data normalised by its rolling mean.
"""
# -- Calculate the average correlation over all freq_range_fft_bins between source and pattern
peaks = correlation_picking(source_file_path,pattern,config_par['freq_range_fft_bins'],
len_fft = config_par["len_fft"],
nr_ffts_per_s = config_par["nr_ffts_per_s"],
pattern_len_s = config_par["pattern_len_sec"]
)
# -- Identify frames of isolated peaks
centre_frames,roll_max_peaks = peak_identification(peaks,
width_in_s = config_par["peak_mute_width"],
width_roll_mean = config_par["width_roll_mean"],
roll_max_peaks_threshold = config_par['peaks_threshold'],
fs = fs,nr_ffts_per_s = config_par["nr_ffts_per_s"],
chunk_len_s = 60,len_fft = config_par['len_fft'],
is_ret_roll_max_peaks = True)
print 'centre_frames',centre_frames
# -- Create range dictionary around isolated peaks
range_dict = intvpck.training_range_times_from_peaktimes(centre_frames,
config_par["range_len_sec"],source_file_path,fs)
return range_dict,roll_max_peaks
