def classify_encapsulated(audio_summary, track_info, pickle_file):
"""Performs a similar operation to classify() and all the above functions in one step
But retrieves fit from pickle, for the GUI
"""
f = open(pickle_file, 'r')
fit = pickle.load(f)
f.close()
rv = {}
#print track_info.keys()
#print track_info[audio_summary]['title'].keys()
rv["speechiness"] = [audio_summary['audio_summary']['speechiness']]
rv["time_sig"] = [audio_summary['audio_summary']['time_signature']]
rv["bpm"] = [audio_summary['audio_summary']['tempo']]
rv["key"] = [audio_summary['audio_summary']['key']]
rv["duration"] = [audio_summary['audio_summary']['duration']]
rv["loudness"] = [audio_summary['audio_summary']['loudness']]
rv["end_of_fade_in"] = [track_info['track']['end_of_fade_in']]
rv["start_of_fade_out"] = [track_info['track']['start_of_fade_out']]
rv["bpm_range"] = [proc.bpm_range(track_info['beats'])]
rv["max_bpm_spike"] = [proc.max_bpm_spike(track_info['beats'])]
try:
rv["num_keys"] = [proc.num_keys(track_info['sections'])]
except:
rv["num_keys"] = [1]
rv["sections"] = [proc.num_sections(track_info['sections'])]
new_df = build_data_frame([rv],["Unknown"])
p = prob_category(new_df,fit)
robjects.globalenv["pred"] = p
edm_prob = robjects.default_ri2py(p.rx(1))[0]
folk_prob = robjects.default_ri2py(p.rx(2))[0]
rap_prob = robjects.default_ri2py(p.rx(3))[0]
cls = classify(new_df,fit)
return [(edm_prob,folk_prob,rap_prob),cls[0]]
def query(name,artist):
"""Prints summary of song and returns information necessary for decision tree"""
keys = ['C', 'C#', 'D', 'Eb', 'E', 'F', 'F#', 'G', 'Ab', 'A', 'Bb']
if name == "" or artist == "":
display_text("No results for that query. \n")
return -1
dict = song_profile(str(name),str(artist))
if dict != "No results for that query. \n":
dict2 = gti.get_json(dict['audio_summary']['analysis_url'])
tempo = dict['audio_summary']['tempo']
time_sig = dict['audio_summary']['time_signature']
speechiness = dict['audio_summary']['speechiness']
artist_name = dict['artist_name']
song = dict['title']
key_not_found = False
try:
key = keys[dict['audio_summary']['key']]
except:
key_not_found = True
key = 'C'
duration = dict['audio_summary']['duration']
loudness = dict['audio_summary']['loudness']
bpm_range = procinp.bpm_range(dict2['beats'])
max_bpm_spike = procinp.max_bpm_spike(dict2['beats'])
num_sections = procinp.num_sections(dict2['sections'])
num_keys_not_found = False
try:
num_keys = procinp.num_keys(dict2['sections'])
except:
num_keys_not_found = True
num_keys = 1
a = "Song information:\n"
b = "\t" + song + ", by " + artist_name + "\n"
c = "\tTime signature: " + str(time_sig)+ "\n"
d = "\tTempo: " + str(tempo)+ " beats per minute\n"
e = "\tSpeechiness: " + str(speechiness)+ "\n"
f = "\tLoudness: " + str(loudness)+ " decibels\n"
if key_not_found:
g = "\tKey: " + 'Unknown'+ "\n"
else:
g = "\tKey: " + str(key)+ "\n"
h = "\tDuration: " + str(duration)+ " seconds\n"
i = "\tTempo Range: "+str(bpm_range)+"\n"
j = "\tMaximum Tempo Spike: "+str(max_bpm_spike)+"\n"
k = "\tNumber of Sections: "+str(num_sections)+"\n"
m = "\n"
if num_keys_not_found:
l = "\tNumber of Keys: "+'Unknown'+"\n"
else:
l = "\tNumber of Keys: "+str(num_keys)+"\n"
display_text(a+b+c+d+e+f+g+h+i+j+k+l+m)
return (dict,dict2)
else:
display_text(dict)
return -1