Question

几周前我刚接触python，并且一直在使用PyCharm软件学习该语言。我希望为项目使用以下代码： https://github.com/eohomegrownapps/markov-music

我安装了所需的库，并在上述文件夹（输入文件夹）中插入了一个midi文件。我正在使用以下midi文件： http://www.bachcentral.com/invent/invent1.mid

运行代码后，“ compositions”文件夹中将不会出现任何内容。谁能告诉我我在做什么错？我将在下面的代码中截图。

预先感谢一百万

import random
import midi
import datetime


class Markov(object):
    """docstring for Markov"""

    def __init__(self, order=2):
        super(Markov, self).__init__()
        self.order = order
        self.chain = {}

    def add(self, key, value):
        if self.chain.has_key(key):
            self.chain[key].append(value)
        else:
            self.chain[key] = [value]

    def load(self, midifile):
        pattern = midi.read_midifile(midifile)
        # track = pattern[1]
        for track in pattern:
            noteslist = []
            curoffset = 0
            for i in track:
                if i.name == "Note On" and i.data[1] != 0:
                    note = (i.data[0], i.data[1], i.tick + curoffset)
                    # note = (i.data[0],i.data[1],i.tick)
                    noteslist.append(note)
                    curoffset = 0
                else:
                    curoffset += i.tick
            if len(noteslist) > self.order:
                for j in range(self.order, len(noteslist)):
                    t = tuple(noteslist[j - self.order:j])
                    print t
                    print noteslist[j]
                    self.add(t, noteslist[j])
            else:
                print "Corpus too short"

    def generate(self, length, filename):
        pattern = midi.Pattern()
        # Instantiate a MIDI Track (contains a list of MIDI events)
        track = midi.Track()
        # Append the track to the pattern
        pattern.append(track)

        tick = 0
        currenttuple = random.choice(self.chain.keys())
        prevnote = False
        for i in range(0, self.order):
            if prevnote != False:
                on = midi.NoteOnEvent(tick=tick, velocity=0, pitch=prevnote)
                track.append(on)
            on = midi.NoteOnEvent(tick=0, velocity=currenttuple[i][1], pitch=currenttuple[i][0])
            track.append(on)
            tick = currenttuple[i][2]
            prevnote = currenttuple[i][0]
        result = random.choice(self.chain[currenttuple])
        # print currenttuple
        for i in range(1, length):
            for j in range(0, self.order):
                if prevnote != False:
                    if tick > 5000:
                        tick = 5000
                    on = midi.NoteOnEvent(tick=tick, velocity=0, pitch=prevnote)
                    track.append(on)
                on = midi.NoteOnEvent(tick=0, velocity=currenttuple[j][1], pitch=currenttuple[j][0])
                track.append(on)
                tick = currenttuple[j][2]
                prevnote = currenttuple[j][0]

            currenttuple = list(currenttuple)
            currenttuple.pop(0)
            currenttuple.append(result)
            currenttuple = tuple(currenttuple)
            if self.chain.has_key(currenttuple):
                result = random.choice(self.chain[currenttuple])
            else:
                result = random.choice(self.chain[random.choice(self.chain.keys())])
        # Add the end of track event, append it to the track
        eot = midi.EndOfTrackEvent(tick=1)
        track.append(eot)
        # Print out the pattern
        print pattern
        # Save the pattern to disk
        midi.write_midifile(filename + ".mid", pattern)


if __name__ == '__main__':
    directory = "compositions"
    musicdir = "input"

    musicdir += "/"
    logname = directory + "/" + "{:%Y-%m-%d-%H:%M:%S}_genmusic".format(datetime.datetime.now())

    m = Markov(3)
    print "Loading music"
    inp = raw_input('Name of midi file to load or g to generate: ')
    while inp != "g":
        try:
            m.load(musicdir + inp)
        except Exception as e:
            print "File not found or corrupt"
        inp = raw_input('Name of midi file to load or g to generate: ')
    print "Done"
    print m.chain
    m.generate(1000, logname)

运行python代码时遇到问题（Markov Chain项目）

0 个答案: