Unable to install

[reActionHank]

tried installing it on 2 separate machines, but I get the same error:

Error in install_keras(method = method, conda = conda, version = keras, : You should call install_keras() only in a fresh R session that has not yet initialized Keras and TensorFlow (this is to avoid DLL in use errors during installation)

Tried a couple of different things, creating new anaconda environments, reinstalling python, installing miniconda, etc.

I can install the keras and tensorflow packages separately w/o much problems, just the autokeras is an issue.

If you have any idea what to do would appreaciate it.

[jcrodriguez1989]

Hi @reActionHank , thanks for your feedback.
I am trying to reproduce your issue, so far, I am guessing you are using Windows.
So, in my Windows 10 VM I am not being able to even install keras.
Have you tried starting an R session and doing

remotes::install_github("rstudio/keras")
library("keras")
install_keras()

Is this working for you?

thanks

[reActionHank]

Hey, thanks for the fast response. I've managed to install keras, but it was difficult. It worked with an older version of Miniconda3 (4.11). This also helped: rstudio/keras#256 Unfortunately, I'm not able to do the same with autokeras.

…

On Sun, Apr 21, 2019 at 5:37 PM Juan C Rodriguez ***@***.***> wrote: Hi @reActionHank <https://github.com/reActionHank> , thanks for your feedback. I am trying to reproduce your issue, so far, I am guessing you are using Windows. So, in my Windows 10 VM I am not being able to even install keras. Have you tried starting an R session and doing remotes::install_github("rstudio/keras") library("keras") install_keras() Is this working for you? thanks — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AH5OVPDLHQHD7QMNG6VWMHTPRR32BANCNFSM4HHKMMNQ> .

[jcrodriguez1989]

@reActionHank I am not being able to reproduce your issue since I am not a Windows user, and my VM needs a lot of libraries installation.
Your Error in install_keras(method = method, ... error is not an autokeras code error. If you can reproduce it by creating a new empty env and try install_keras() then you might post it as a keras issue, this would be helpful for both autokeras and keras R libraries :) .
However, as a workaround for autokeras, you could manually install the Python Auto-Keras library.
You could try activating your keras working environment (remember it must have Python 3.6) and do from prompt:

conda install pytorch torchvision -c pytorch
pip install autokeras

[jcrodriguez1989]

@reActionHank I think what you noticed, might be an issue. Could you please reinstall autokeras remotes::install_github("jcrodriguez1989/autokeras") and re try doing install_autokeras()?
thanks!

In install_autokeras() was using reticulate::py_config() to get current Python version, and using this function initializes Python, and thus when keras::install_keras() calls reticulate::py_available() it returns TRUE and then the error you noted

> reticulate::py_available()
[1] FALSE
> reticulate::py_config()$version
[1] "3.6"
> reticulate::py_available()
[1] TRUE

[reActionHank]

Hey, this is the error I got after trying the new install:

install_autokeras()

Remove all packages in environment C:\Users\Desktop\MINICO~1\envs\r-tensorflow: ## Package Plan ## environment location: C:\Users\Desktop\MINICO~1\envs\r-tensorflow The following packages will be REMOVED: _tflow_select: 2.3.0-mkl absl-py: 0.7.1-py37_0 conda-forge asn1crypto: 0.24.0-py37_1003 conda-forge astor: 0.7.1-py_0 conda-forge blas: 1.0-mkl ca-certificates: 2019.3.9-hecc5488_0 conda-forge certifi: 2019.3.9-py37_0 conda-forge cffi: 1.12.3-py37hb32ad35_0 conda-forge chardet: 3.0.4-py37_1003 conda-forge cryptography: 2.6.1-py37hb32ad35_0 conda-forge freetype: 2.10.0-h5db478b_0 conda-forge gast: 0.2.2-py_0 conda-forge grpcio: 1.16.1-py37h351948d_1 h5py: 2.9.0-nompi_py37h3cb27cb_1102 conda-forge hdf5: 1.10.4-nompi_hcc15c50_1106 conda-forge icc_rt: 2019.0.0-h0cc432a_1 idna: 2.8-py37_1000 conda-forge intel-openmp: 2019.3-203 jpeg: 9c-hfa6e2cd_1001 conda-forge keras: 2.2.4-0 keras-applications: 1.0.7-py_0 conda-forge keras-base: 2.2.4-py37_0 keras-preprocessing: 1.0.9-py_0 conda-forge libblas: 3.8.0-5_mkl conda-forge libcblas: 3.8.0-5_mkl conda-forge liblapack: 3.8.0-5_mkl conda-forge libmklml: 2019.0.3-0 libpng: 1.6.37-h7602738_0 conda-forge libprotobuf: 3.7.1-h1a1b453_0 conda-forge libtiff: 4.0.10-h36446d0_1001 conda-forge markdown: 2.6.11-py_0 conda-forge mkl: 2019.3-203 numpy: 1.16.2-py37h8078771_1 conda-forge olefile: 0.46-py_0 conda-forge openssl: 1.1.1b-hfa6e2cd_2 conda-forge pillow: 6.0.0-py37h9a613e6_0 conda-forge pip: 19.0.3-py37_0 protobuf: 3.7.1-py37he025d50_0 conda-forge pycparser: 2.19-py37_1 conda-forge pyopenssl: 19.0.0-py37_0 conda-forge pyreadline: 2.1-py37_1000 conda-forge pysocks: 1.6.8-py37_1002 conda-forge python: 3.7.3-h8c8aaf0_0 pyyaml: 5.1-py37hfa6e2cd_0 conda-forge requests: 2.21.0-py37_1000 conda-forge scipy: 1.2.1-py37h29ff71c_0 setuptools: 41.0.0-py37_0 six: 1.12.0-py37_1000 conda-forge sqlite: 3.27.2-he774522_0 tensorboard: 1.13.1-py37_0 conda-forge tensorflow: 1.13.1-mkl_py37h9463c59_0 tensorflow-base: 1.13.1-mkl_py37hcaf7020_0 tensorflow-estimator: 1.13.0-py37h39e3cac_0 conda-forge tensorflow-hub: 0.4.0-py_0 conda-forge tensorflow-probability: 0.5.0-py_1 conda-forge termcolor: 1.1.0-py_2 conda-forge tk: 8.6.9-hfa6e2cd_1001 conda-forge urllib3: 1.24.2-py37_0 conda-forge vc: 14.1-h0510ff6_4 vs2015_runtime: 14.15.26706-h3a45250_0 werkzeug: 0.15.2-py_0 conda-forge wheel: 0.33.1-py37_0 win_inet_pton: 1.1.0-py37_0 conda-forge wincertstore: 0.2-py37_0 yaml: 0.1.7-hfa6e2cd_1001 conda-forge zlib: 1.2.11-h2fa13f4_1004 conda-forge Creating r-tensorflow conda environment for TensorFlow installation... Solving environment: ...working... done ==> WARNING: A newer version of conda exists. <== current version: 4.5.11 latest version: 4.6.14 Please update conda by running $ conda update -n base -c defaults conda sqlite-3.28.0 | 945 KB | ########## | 100% certifi-2019.3.9 | 155 KB | ########## | 100% ## Package Plan ## environment location: C:\Users\Desktop\MINICO~1\envs\r-tensorflow added / updated specs: - python=3.7 The following packages will be downloaded: package | build ---------------------------|----------------- sqlite-3.28.0 | he774522_0 945 KB certifi-2019.3.9 | py37_0 155 KB ------------------------------------------------------------ Total: 1.1 MB The following NEW packages will be INSTALLED: ca-certificates: 2019.1.23-0 certifi: 2019.3.9-py37_0 openssl: 1.1.1b-he774522_1 pip: 19.0.3-py37_0 python: 3.7.3-h8c8aaf0_0 setuptools: 41.0.0-py37_0 sqlite: 3.28.0-he774522_0 vc: 14.1-h0510ff6_4 vs2015_runtime: 14.15.26706-h3a45250_0 wheel: 0.33.1-py37_0 wincertstore: 0.2-py37_0 Downloading and Extracting Packages Preparing transaction: ...working... done Verifying transaction: ...working... done Executing transaction: ...working... done # # To activate this environment, use: # > activate r-tensorflow # # To deactivate an active environment, use: # > deactivate # # * for power-users using bash, you must source # Solving environment: ...working... failed PackagesNotFoundError: The following packages are not available from current channels: - autokeras==0.3.7 Current channels: - https://conda.anaconda.org/conda-forge/win-64 - https://conda.anaconda.org/conda-forge/noarch - https://repo.anaconda.com/pkgs/main/win-64 - https://repo.anaconda.com/pkgs/main/noarch - https://repo.anaconda.com/pkgs/free/win-64 - https://repo.anaconda.com/pkgs/free/noarch - https://repo.anaconda.com/pkgs/r/win-64 - https://repo.anaconda.com/pkgs/r/noarch - https://repo.anaconda.com/pkgs/pro/win-64 - https://repo.anaconda.com/pkgs/pro/noarch - https://repo.anaconda.com/pkgs/msys2/win-64 - https://repo.anaconda.com/pkgs/msys2/noarch To search for alternate channels that may provide the conda package you're looking for, navigate to https://anaconda.org and use the search bar at the top of the page. Error: Error 1 occurred installing packages into conda environment r-tensorflow

…

On Thu, Apr 25, 2019 at 4:59 AM Juan C Rodriguez ***@***.***> wrote: @reActionHank <https://github.com/reActionHank> I think what you noticed, might be an issue. Could you please reinstall autokeras remotes::install_github("jcrodriguez1989/autokeras") and re try doing install_autokeras()? thanks! In install_autokeras() was using reticulate::py_config() to get current Python version, and using this function initializes Python, and thus when keras::install_keras() calls reticulate::py_available() it returns TRUE and then the error you noted > reticulate::py_available() [1] FALSE> reticulate::py_config()$version [1] "3.6"> reticulate::py_available() [1] TRUE — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AH5OVPGJLJMKDIZOOSPG7A3PSEF73ANCNFSM4HHKMMNQ> .

[jcrodriguez1989]

thanks!
So, the bug is over. Now it is working at it should, and it is saying that the conda env is going to use Python 3.7, which is not supported by autokeras. And there it appears the new issue you are having.
I have been checking this, and Python 3.7 is being used because tensorflow R library uses this Python version if using default tensorflow version (1.3).
However, if you specify tensorflow 1.2. It will have other problems related to pytorch and etc.
To summarise, I would recommend you installing Python autokeras manually in your conda env.

[loheden]

@jcrodriguez1989 , how do you install autokeras manually from source? I am not able to install it on MAC OS M1.