We Keep Coding

DDSP colab demo import error / ModuleNotFoundError: No module named 'crepe'

I tried to use the demo of the DDSP model developed by Google Research in colab, but there was an error during the import stage, so I couldn't import it.
This problem didn't happen until recently, but suddenly an error is occurring.
`
##title #Install and Import
##markdown Install ddsp, define some helper functions, and download the model. This transfers a lot of data and _should take a minute or two_.
print('Installing from pip package...')
!pip install -qU ddsp==1.6.5 "hmmlearn<=0.2.7"
# Ignore a bunch of deprecation warnings
import warnings
warnings.filterwarnings("ignore")
import copy
import os
import time
import crepe
import ddsp
import ddsp.training
from ddsp.colab.colab_utils import (
auto_tune, get_tuning_factor, download,
play, record, specplot, upload,
DEFAULT_SAMPLE_RATE)
from ddsp.training.postprocessing import (
detect_notes, fit_quantile_transform
)
import gin
from google.colab import files
import librosa
import matplotlib.pyplot as plt
import numpy as np
import pickle
import tensorflow.compat.v2 as tf
import tensorflow_datasets as tfds
# Helper Functions
sample_rate = DEFAULT_SAMPLE_RATE # 16000
print('Done!')
`
It seems to be a problem with the version, so I tried to upgrade the pip and install the pip before importing the module according to the guide, but it still hasn't been solved.
Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.
WARNING: etils 0.9.0 does not provide the extra 'epath-no-tf'
A similar warning will continue to be displayed below...
`
---------------------------------------------------------------------------
ModuleNotFoundError Traceback (most recent call last)
<ipython-input-2-fb97381ec3fd> in <module>
13 import time
14
---> 15 import crepe
16 import ddsp
17 import ddsp.training
ModuleNotFoundError: No module named 'crepe'
---------------------------------------------------------------------------
NOTE: If your import is failing due to a missing package, you can
manually install dependencies using either !pip or !apt.
To view examples of installing some common dependencies, click the
"Open Examples" button below.
---------------------------------------------------------------------------
`
This error is output if I stop importing.

ModuleNotFoundError: No module named 'pycaret.time_series'

I've seen similar questions, but no solutions for this problem.
I'm using Google Colab:
!pip install pycaret
And this:
import time
import numpy as np
import pandas as pd
from pycaret.datasets import get_data
#from pycaret.time_series import TSForecastingExperiment
from pycaret.time_series import *
The last line returns the error: ModuleNotFoundError: No module named 'pycaret.time_series'
Does anyone know how to fix this?
Thanks in advance!

Please follow the instructions here: https://github.com/pycaret/pycaret#-pycaret-time-series-module-beta
Specifically, for the time being, the Time Series module needs to be installed separately in its own environment (i.e. not together with the main pycaret package), using:
!pip install pycaret-ts-alpha

Is there a way to import an older version of tensor flow to use in python?

I keep getting this error when running my python 3.8 code:
AttributeError: module 'tensorflow._api.v2.train' has no attribute 'SessionRunHook'
I previously read that this is because the tensorflow version is not correct. How do I use a different version of tensorflow?
I have tried:
import tensorflow as tf
tf.compat.v1.disable_v2_behavior()
print(tf.__version__)
but it still shows that I am using tensorflow version 2.5.0.
I have also tried
from __future__ import unicode_literals
from __future__ import print_function
from __future__ import division
from __future__ import absolute_import
import argparse
import os
import pandas as pd
os.environ["CUDA_VISIBLE_DEVICES"] = '0'
import tensorflow.compat.v1 as tf
tf.compat.v1.disable_v2_behavior()
print(tf.__version__)
import numpy as np
from dltk.networks.regression_classification.resnet import resnet_3d
from dltk.io.abstract_reader import Reader
The attribute error seems to appear at the last line when I try to import dltk.io.abstract_reader.

You can uninstall the package and install an older version. The grammar is like pip install tensorflow==<version>.
If you would like to keep the two versions at the same time, I'm afraid you should use anaconda.

How to Import Image from PILLOW In Jupyter?

I am trying to import Image from PILLOW.
from PILLOW import Image
im = Image.open("bride.jpg")
im.rotate(45).show()
Undefined function 'from' for input arguments of type 'char'.
>> import PIL.Image
im = Image.open("bride.jpg")
im.rotate(45).show()
Error using import
Import argument 'PIL.Image' cannot be found or cannot be imported.
Undefined function 'from' for input arguments of type 'char'.
I keep getting the same error. Am I using the wrong import statement?
EDIT:
I noticed something else. when i browse through my C:/ drive and look inside the folders where Pip, and PIL ,etc is. There is nothing inside them. Even when i copy a file with stuff to that area, and open that new file, i cant see any of the stuff inside. When i copy it though, and the green bar goes it shows data being transferred.
The Code:
https://hub.gke.mybinder.org/user/alexmill-website_notebooks-fcqkeh7a/notebooks/fourier-spinning-circles.ipynb#
Example:
enter code here
from PIL import Image, ImageEnhance
# Main imports/Initialization
%matplotlib inline
import matplotlib
from matplotlib import animation, rc
import matplotlib.pyplot as plt
from matplotlib.pyplot import imshow
from IPython.display import HTML
import numpy as np
from PIL import Image, ImageEnhance
import requests
from io import BytesIO
from copy import deepcopy
from scipy.spatial import distance
from scipy.interpolate import UnivariateSpline
from copy import deepcopy
​
# Default figure size in notebook
matplotlib.rcParams['figure.figsize'] = (6,6)
matplotlib.rcParams['image.aspect'] = 'equal'
---------------------------------------------------------------------------
ModuleNotFoundError Traceback (most recent call last)
<ipython-input-16-7c75157a7886> in <module>
7 from IPython.display import HTML
8 import numpy as np
----> 9 from PIL import Image, ImageEnhance
10 import requests
11 from io import BytesIO
ModuleNotFoundError: No module named 'PIL'

The reason your binder doesn't work is because your requirements.txt looks like this:
matplotlib==2.0.2
requests==2.20.0
Pillow==5.0.0
numpy==1.13.0
scipy==0.19.0
pandas==0.20.2
scikit-learn==0.18.1
tzlocal-1.5.1
This line:
tzlocal-1.5.1
Probably breaks the binder dependency install. Another thing is that these are very specific versions that might generate too specific dependencies (ref. here). I recommend changing it to this:
matplotlib
requests
pillow
numpy
scipy
pandas
scikit-learn
tzlocal
At least your binder started working when I did that.
Locally you will just have to run pip install -r requirements.txt with an admin command prompt and then your local Jupyter installation should automatically work, at least it did when I cloned your notebook to a Windows VM.

Scikit-image: cannot import name 'label'

I have some scripts that use skimage.measure.label. On my old laptop (Debian 7, Python 2) these scripts worked perfectly. Recently, when I got a new laptop, I moved to Debian 8 and Python 3. Now these scripts cannot import skimage.measure.label:
File "image_converter.py", line 8, in <module>
from skimage.measure import label
ImportError: cannot import name 'label'
All of the other imports are working fine:
from skimage import data
from skimage.filter import threshold_otsu
from skimage.segmentation import clear_border
from skimage.morphology import closing, square
from skimage.measure import regionprops
from skimage.color import label2rgb
import skimage.io as ski_io
pip3 list says:
(...)
scikit-image (0.9.3)
(...)
Python version is 3.4.3.
What has happened there? Has scikit-image (re-)moved measure.label?
In the official documentation it is still listed: skimage.measure.label

There is indeed no function label in module measure for release 0.9 (see http://scikit-image.org/docs/0.9.x/api/skimage.measure.html).
Basically, you should update your version of skimage. The most recent version is 0.12.3. Python 3.4 is officially tested against, so you shouldn't expect any troubles with compatibility.