Closed lenthomas closed 5 years ago
goldingn
commented
6 years ago It's possible anaconda was installed, or runs with one account (e.g. it might have asked for an administrator password during installation) but R / RStudio is being run with another. If so, you might be able to fix it by running R / RStudio as administrator and trying the conda installation of TF and TFP again.
rdinnager
commented
5 years ago @goldingn I can confirm that running Rstudio as administrator does not solve this issue. I get the exact same error message as @lenthomas running as administrator or not.. in Windows 10.. on two separate machines.
flyaflya
commented
5 years ago Sorry to be late to this discussion, but I now have some experience teaching greta to MBA Business Analytics students. I put together these instructions for install: https://bit.ly/2AXb4Yg
28 students used these instructions successfully on a mix of Windows and Mac machines. Maybe it can help you?
lenthomas
commented
5 years ago thanks for the suggestion @flyaflya. I followed the instructions (below); everything seemed to go well, but when I called
library(greta)
I got the same error message as before
`greta requires TensorFlow (>=1.10.0) and Tensorflow Probability (>=0.3.0), but TensorFlow Probability isn't installed. Use:
install_tensorflow(method = "conda")
reticulate::conda_install("r-tensorflow", "tensorflow-probability", pip = TRUE)to install the latest version.`
The instructions I followed were:
conda create -n r-tensorflow python=3.5 pip
conda activate r-tensorflow
pip install --no-dependencies tensorflow-probability==0.4.0
conda install tensorflow=1.11
pip install h5py pyyaml requests Pillow scipyinstall.packages("greta")
library(greta)Failed on the last step :(
flyaflya
commented
5 years ago I have wrestled with this exact issue. The key marker of success is that within the r-tensorflow environment (i.e. reached via activate r-tensorflow from the anaconda prompt) to have
conda list show installations of both tensorflow and tensorflow-probability. Knowing this can help debug. Also, do NOT run the reticulate::conda_install(... as that will undo everything you just did.
I have three suggestions to try to get it working in order of effort:
Error message suggests you are only missing tensorflow-probability. Just add tensorflow-probability by rerunning these two lines from the anaconda prompt:
conda activate r-tensorflow
pip install --no-dependencies tensorflow-probability==0.4.0Go through steps again with a clean slate. Run these two lines to delete the r-tensorflow environment and go through the steps you did previously.:
deactivate
conda env remove --name r-tensorflowIn above, use source deactivate if on a Mac.
Full debug: Do step 2, but run conda list after each step. Record what is happening with the tensorflow and tensorflow-probability packages. Try to get to a state where they are both decent versions ... (note: Python will switch up versions for compatibility). My current state is shown below:
If you need to get to step 3, try to figure out which step is wiping out tensorflow-probability and we can try to debug from there.
lenthomas
commented
5 years ago Thanks so much for your help with this. I went back to the beginning, deactivating and then removing r-tensorflow, and then worked through steps 2 and 3, running conda list after each one. Here's what happened - bottom line is same problem as before.
lenthomas
commented
5 years ago Note that tensorflow and tensorflow-probability are both listed by conda list.
flyaflya
commented
5 years ago Hmmm. Unfortunately, I am starting to swim outside of my depth here and maybe we can get more help on this from others. Seems the key issue now might be either 1) R is not finding your r-tensorflow environment (maybe a Python path?) or 2) something with that eigen... build in your screenshot (I am not familiar with this... Python is still quite foreign to me). The beauty of greta is you do not have to leave R, but these Python config issues sure are frustrating.
These lines run in R might fix the issue by using the right environment or help with debugging:
library(tensorflow)
use_condaenv("r-tensorflow", required = TRUE)
tf_config() ## should match conda list
reticulate::py_module_available("tensorflow") ## should return true
reticulate::py_module_available("tensorflow_probability") ## should return true
library("greta")Thanks. Yup, does seem to be a path issue
Seemingly I have python in my path twice!
lenthomas
commented
5 years ago Weirdly, if I call tf_config() before use_condaenv, I get further - although it apparently still can't find tensorflow_probability, and I really don't understand why calling tf_config first makes any difference!
lenthomas
commented
5 years ago Perhaps a problem is that I'm using Python 3.6 and Tensorflow 1.10.0. I'm not quite sure where they are coming from, since I thought in following your instructions I'd end up with Python 3.6 and Tensorflow 1.11. Urgh!
rdinnager
commented
5 years ago Thanks so much for your help @flyaflya . I eventually got this working, though I had to use a slightly different sequence. This is possibly because I am trying to install the development version of greta which would seem to require tensorflow-probability > 0.5.0. So I just replaced the lines
pip install --no-dependencies tensorflow-probability==0.4.0
conda install tensorflow=1.11with:
pip install --no-dependencies tensorflow-probability==0.5.0
conda install tensorflow=1.12Kept everything else the same and it worked!
One thing I found helpful while I was debugging was to try reticulate::import("tensorflow_probability"). I found that reticulate::py_module_available would return FALSE not only if a module was not installed but also if it won't load. reticulate::import would then return a more informative error message if the problem was that tensorflow-probability just wasn't loading. It wasn't loading on mine at one point because of a mismatch in the versions of my tensorflow-probability and my tensorflow modules. Perhaps it might help diagnose your problems @lenthomas ?
flyaflya
commented
5 years ago Glad the instructions were helpful for you @rdinnager. Thanks for saying so.
@lenthomas the tf_config() line makes it clear to me that R is finding tensorflow in a different location than the one you set-up using mini-conda (the paths and the tf versions are different). I am sure there is an elegant way to fix this issue, but I only know the brute force way. This is to uninstall all Python installations and start fresh. It seems that you might have both Anaconda and Mini-Conda installed which seems to be a problem: https://stackoverflow.com/questions/51585854/install-both-anaconda-and-miniconda
Since, this is really either: 1) R finding the a workable tensorflow environment with tfp or 2) possibly a Python configuration issue and not a bug with greta, I would recommend either using https://stackoverflow.com/ or the greta forum at https://forum.greta-stats.org/ to dig deeper if the brute force reinstall of Python does not fix things.
@goldingn - I would consider closing this issue as it is not directly a greta issue. However, if there were a way to make greta more robust to these types of Python configuration issues, that would be great enhancement... but even that might be something for the fine people at https://github.com/rstudio/tensorflow/ .
eteppo
commented
5 years ago Thank you. My issue was solved like below.
What greta needs and what it's detecting?
library(greta)
tensorflow::use_condaenv("r-tensorflow")
tensorflow::tf_config()
reticulate::py_module_available("tensorflow")
reticulate::py_module_available("tensorflow_probability")
reticulate::import("tensorflow_probability")Fix install manually via shell (Ubuntu):
source activate r-tensorflow
pip install tensorflow==<CORRECT VERSION>
pip install tensorflow-probablity==<CORRECT VERSION>
reboot
tiboloic
commented
5 years ago I there, I had exactly the same problem than @lenthomas. I manage to solve it using the instructions provided by @flyaflya. Specifically I had to remove r-tensorflow and recreate it in Anaconda's command line interface. Thanks all
goldingn
commented
5 years ago I'm going to close this, as I think the latest version of r tensorflow (which greta now depends on) does a good job of installing these correctly. Please open a new issue if you have similar problems with the greta 0.3.1 though
Sorry to be a bother, but: different computer, different install issue. Full transcript below. Appears to be a permission issue, which is strange since Anacaonda installed itself into a user directory where there should be no such issues. Any suggestions on this one? Thanks!
Win 10; R 3.5.1.
Transcript starts:
R version 3.5.1 (2018-07-02) -- "Feather Spray" Copyright (C) 2018 The R Foundation for Statistical Computing Platform: x86_64-w64-mingw32/x64 (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY. You are welcome to redistribute it under certain conditions. Type 'license()' or 'licence()' for distribution details.
R is a collaborative project with many contributors. Type 'contributors()' for more information and 'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or 'help.start()' for an HTML browser interface to help. Type 'q()' to quit R.
greta requires TensorFlow (>=1.10.0) and Tensorflow Probability (>=0.5.0), but you have TensorFlow version 1.4.0 and TensorFlow Probability isn't installed. Use:
to install the latest version.
Attaching package: ‘greta’
The following objects are masked from ‘package:stats’:
The following objects are masked from ‘package:base’:
Package plan for package removal in environment C:\Users\len\ANACON~1\envs\r-tensorflow:
The following packages will be REMOVED:
Unlinking packages ... [ COMPLETE ]|##################################################| 100% Creating r-tensorflow conda environment for TensorFlow installation... Fetching package metadata ....... Solving package specifications: ..........
Package plan for installation in environment C:\Users\len\ANACON~1\envs\r-tensorflow:
The following NEW packages will be INSTALLED:
Linking packages ... [ COMPLETE ]|##################################################| 100% #
To activate this environment, use:
$ source activate r-tensorflow
#
To deactivate this environment, use:
$ source deactivate
# Fetching package metadata ......... Solving package specifications: ..........
Package plan for installation in environment C:\Users\len\ANACON~1\envs\r-tensorflow:
The following packages will be downloaded:
The following NEW packages will be INSTALLED:
The following packages will be UPDATED:
Fetching packages ... blas-1.0-mkl.t 100% |###############################| Time: 0:00:00 386.44 kB/s ca-certificate 100% |###############################| Time: 0:00:00 440.73 kB/s vs2015win-64- 100% |###############################| Time: 0:00:00 291.07 kB/s libgpuarray-0. 100% |###############################| Time: 0:00:00 592.26 kB/s openssl-1.0.2o 100% |###############################| Time: 0:00:10 556.72 kB/s hdf5-1.10.3-he 100% |###############################| Time: 0:00:31 1.16 MB/s jpeg-9c-hfa6e2 100% |###############################| Time: 0:00:00 598.39 kB/s libprotobuf-3. 100% |###############################| Time: 0:00:09 1.08 MB/s tk-8.6.8-vc14 100% |###############################| Time: 0:00:04 881.52 kB/s asn1crypto-0.2 100% |###############################| Time: 0:00:00 359.19 kB/s backports-1.0- 100% |###############################| Time: 0:00:00 239.74 kB/s certifi-2018.1 100% |###############################| Time: 0:00:01 77.15 kB/s chardet-3.0.4- 100% |###############################| Time: 0:00:00 438.79 kB/s cryptography-v 100% |###############################| Time: 0:00:22 1.41 MB/s freetype-2.9.1 100% |###############################| Time: 0:00:00 709.39 kB/s idna-2.7-py36_ 100% |###############################| Time: 0:00:00 161.22 kB/s markdown-2.6.1 100% |###############################| Time: 0:00:00 287.98 kB/s markupsafe-1.1 100% |###############################| Time: 0:00:00 241.91 kB/s olefile-0.46-p 100% |###############################| Time: 0:00:00 313.22 kB/s pycparser-2.19 100% |###############################| Time: 0:00:00 418.33 kB/s pyreadline-2.1 100% |###############################| Time: 0:00:00 434.41 kB/s pyyaml-3.13-py 100% |###############################| Time: 0:00:00 451.22 kB/s six-1.11.0-py3 100% |###############################| Time: 0:00:00 177.27 kB/s webencodings-0 100% |###############################| Time: 0:00:00 743.82 kB/s win_inetpton- 100% |###############################| Time: 0:00:00 349.56 kB/s absl-py-0.6.1- 100% |###############################| Time: 0:00:00 294.68 kB/s backports.weak 100% |###############################| Time: 0:00:00 468.25 kB/s cffi-1.11.5-py 100% |###############################| Time: 0:00:00 418.72 kB/s h5py-2.8.0-py3 100% |###############################| Time: 0:00:00 856.79 kB/s mako-1.0.7-py 100% |###############################| Time: 0:00:00 289.54 kB/s pillow-5.3.0-p 100% |###############################| Time: 0:00:00 920.52 kB/s pysocks-1.6.8- 100% |###############################| Time: 0:00:00 220.90 kB/s cryptography-2 100% |###############################| Time: 0:00:00 696.47 kB/s protobuf-3.5.2 100% |###############################| Time: 0:00:01 426.18 kB/s pygpu-0.7.6-py 100% |###############################| Time: 0:00:01 355.05 kB/s pyopenssl-18.0 100% |###############################| Time: 0:00:00 355.49 kB/s tensorboard-1. 100% |###############################| Time: 0:00:02 1.32 MB/s theano-1.0.3-p 100% |###############################| Time: 0:00:03 1.22 MB/s tensorflow-1.5 100% |###############################| Time: 0:00:25 1.10 MB/s urllib3-1.23-p 100% |###############################| Time: 0:00:00 459.37 kB/s requests-2.20. 100% |###############################| Time: 0:00:00 377.37 kB/s tensorflow-hub 100% |###############################| Time: 0:00:00 346.53 kB/s keras-applicat 100% |###############################| Time: 0:00:00 264.23 kB/s keras-2.2.4-py 100% |###############################| Time: 0:00:00 667.43 kB/s keras-preproce 100% |###############################| Time: 0:00:00 230.88 kB/s Extracting packages ... [markupsafe ]|###Using Anaconda Cloud api site https://api.anaconda.org INFO:progress.update:('jpeg', 9) INFO:progress.update:('libpng', 10) INFO:progress.update:('libprotobuf', 11) INFO:progress.update:('tk', 12) INFO:progress.update:('asn1crypto', 13) INFO:progress.update:('backports', 14) [ COMPLETE ]|##################################################| 100% Unlinking packages ... [ COMPLETE ]|##################################################| 100% Linking packages ... [certifi ]|############# | INFO:progress.update:('certifi', 15) INFO:progress.update:('chardet', 16) INFO:progress.update:('cryptography-vectors', 17) INFO:progress.update:('freetype', 18) INFO:progress.update:('idna', 19) INFO:progress.update:('libtiff', 20) INFO:progress.update:('markdown', 21) INFO:progress.update:('markupsafe', 22) INFO:progress.update:('numpy', 23) INFO:progress.update:('olefile', 24) INFO:progress.update:('pycparser', 25) INFO:progress.update:('pyreadline', 26) INFO:progress.update:('pyyaml', 27) INFO:progress.update:('six', 28) INFO:progress.update:('webencodings', 29) INFO:progress.update:('werkzeug', 30) INFO:progress.update:('win_inet_pton', 31) INFO:progress.update:('absl-py', 32) INFO:progress.update:('backports.weakref', 33) INFO:progress.update:('cffi', 34) INFO:progress.update:('h5py', 35) INFO:progress.update:('html5lib', 36) INFO:progress.update:('mako', 37) INFO:progress.update:('pillow', 38) INFO:progress.update:('pysocks', 39) INFO:progress.update:('scipy', 40) INFO:progress.update:('bleach', 41) INFO:progress.update:('cryptography', 42) INFO:progress.update:('protobuf', 43) INFO:progress.update:('pygpu', 44) INFO:progress.update:('pyopenssl', 45) INFO:progress.update:('tensorboard', 46) INFO:progress.update:('theano', 47) INFO:progress.update:('tensorflow', 48) WARNING:conda.install:file already exists: 'C:\Users\len\ANACON~1\envs\r-tensorflow\Scripts/tensorboard.exe' INFO:progress.update:('urllib3', 49) INFO:progress.update:('requests', 50) INFO:progress.update:('tensorflow-hub', 51) INFO:progress.update:('keras-applications', 52) INFO:progress.update:('keras', 53) INFO:progress.update:('keras-preprocessing', 54) INFO:progress.stop:None [ COMPLETE ]|##################################################| 100%
Installation complete.
Restarting R session...
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>python C:\Users\len\ANACON~1\envs\r-tensorflow\etc\keras\load_config.py 1>temp.txt
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set /p KERAS_BACKEND= 0<temp.txt
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>del temp.txt
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>python -c "import keras" 1>nul 2>&1
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if errorlevel 1 ( ver 1>nul
set "KERAS_BACKEND=theano"
python -c "import keras" 1>nul 2>&1 )
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET DISTUTILS_USE_SDK=1
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET MSSdk=1
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET platform=
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>IF /I [AMD64] == [amd64] set "platform=true"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>IF /I [] == [amd64] set "platform=true"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if defined platform (set "VSREGKEY=HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\14.0" ) ELSE (set "VSREGKEY=HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\14.0" )
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for /F "skip=2 tokens=2,*" %A in ('reg query "HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\14.0" /v InstallDir') do SET "VSINSTALLDIR=%B"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET "VSINSTALLDIR=C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\" == "" (set "VSINSTALLDIR=C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\Tools\" )
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\" == "" ( ECHO "Did not find VS in registry or in VS140COMNTOOLS env var - exiting"
exit 1 )
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>echo "Found VS2014 at C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\" "Found VS2014 at C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET "VS_VERSION=14.0"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET "VS_MAJOR=14"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET "VS_YEAR=2015"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "MSYS2_ARG_CONV_EXCL=/AI;/AL;/OUT;/out"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "MSYS2_ENV_CONV_EXCL=CL"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "PY_VCRUNTIME_REDIST=\vcruntime140.dll"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "CFLAGS= -MD -GL"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "CXXFLAGS= -MD -GL"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>set "LDFLAGS_SHARED= -LTCG ucrt.lib"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>IF /I [] == [win-64] (set "folder=x64" ) else (set "folder=x86" )
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for /F "tokens=" %I in ('dir "C:\Program Files (x86)\Windows Kits*1" /B /O:N') do for %A in (%~I) do if "%A" == "8.1" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for %A in (10) do if "%A" == "8.1" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "10" == "8.1" set win=10
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for %A in (8.1) do if "%A" == "8.1" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "8.1" == "8.1" set win=8.1
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for /F "tokens=" %I in ('dir "C:\Program Files (x86)\Windows Kits*1" /B /O:N') do for %A in (%~I) do if "%A" == "10" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for %A in (10) do if "%A" == "10" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "10" == "10" set win=10
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>for %A in (8.1) do if "%A" == "10" set win=%A
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "8.1" == "10" set win=8.1
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>setlocal enabledelayedexpansion
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>if "10" == "10" ( for /F "tokens=" %I in ('dir "C:\Program Files (x86)\Windows Kits\10\bin\10" /B /O:N') do for %A in (%~I) do set last=%A
set "sdk_bin_path=C:\Program Files (x86)\Windows Kits\10\bin!last!\x86" ) else (set "sdk_bin_path=C:\Program Files (x86)\Windows Kits\8.1\bin\x86" ) File Not Found
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>endlocal & set "PATH=C:\Users\len\ANACON~1\envs\r-tensorflow;C:\Users\len\ANACON~1\envs\r-tensorflow\Library\mingw-w64\bin;C:\Users\len\ANACON~1\envs\r-tensorflow\Library\usr\bin;C:\Users\len\ANACON~1\envs\r-tensorflow\Library\bin;C:\Users\len\ANACON~1\envs\r-tensorflow\Scripts;C:\Users\len\Anaconda2\envs\r-tensorflow;C:\Users\len\Anaconda2\envs\r-tensorflow\Scripts;C:\Users\len\Anaconda2\envs\r-tensorflow\Library\bin;C:\Program Files\R\R-3.5.1\bin\x64;C:\Program Files (x86)\Common Files\Oracle\Java\javapath;C:\ProgramData\Oracle\Java\javapath;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\Program Files (x86)\Windows Kits\8.1\Windows Performance Toolkit\;C:\Program Files (x86)\Windows Live\Shared;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\Program Files\Git\cmd;c:\x\Rtools\bin;c:\x\Rtools\gcc-4.6.3\bin;C:\Program Files\nodejs\;C:\Program Files (x86)\Pandoc\;C:\WINDOWS\System32\OpenSSH\;C:\X\MiKTeX 2.9\miktex\bin\x64\;C:\Users\len\Anaconda2;C:\Users\len\Anaconda2\Scripts;C:\Users\len\Anaconda2\Library\bin;C:\Users\len\AppData\Local\Microsoft\WindowsApps;C:\Program Files\Microsoft VS Code\bin;C:\Users\len\AppData\Roaming\npm;C:\Program Files (x86)\Windows Kits\10\bin\x86"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>SET "CMAKE_GENERATOR=Visual Studio 14 2015 Win64"
(r-tensorflow) C:\Users\len\ANACON~1\envs\r-tensorflow\etc\conda\activate.d>CALL "C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE....\VC\vcvarsall.bat" amd64
Collecting tensorflow-probability Downloading https://files.pythonhosted.org/packages/a1/ca/6f213618b5f7d0bf6139e6ec928d412a5ca14e4776adfd41a59c74a34021/tensorflow_probability-0.5.0-py2.py3-none-any.whl (680kB) Collecting numpy>=1.13.3 (from tensorflow-probability) Downloading https://files.pythonhosted.org/packages/51/70/7096a735b27359dbc0c380b23b9c9bd05fea62233f95849c43a6b02c5f40/numpy-1.15.4-cp36-none-win_amd64.whl (13.5MB) Collecting six>=1.10.0 (from tensorflow-probability) Downloading https://files.pythonhosted.org/packages/67/4b/141a581104b1f6397bfa78ac9d43d8ad29a7ca43ea90a2d863fe3056e86a/six-1.11.0-py2.py3-none-any.whl Installing collected packages: numpy, six, tensorflow-probability Exception: Traceback (most recent call last): File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\basecommand.py", line 215, in main status = self.run(options, args) File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\commands\install.py", line 342, in run prefix=options.prefix_path, File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\req\req_set.py", line 784, in install **kwargs File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\req\req_install.py", line 851, in install self.move_wheel_files(self.source_dir, root=root, prefix=prefix) File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\req\req_install.py", line 1064, in move_wheel_files isolated=self.isolated, File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\wheel.py", line 345, in move_wheel_files clobber(source, lib_dir, True) File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\site-packages\pip\wheel.py", line 323, in clobber shutil.copyfile(srcfile, destfile) File "C:\Users\len\ANACON~1\envs\r-tensorflow\lib\shutil.py", line 121, in copyfile with open(dst, 'wb') as fdst: PermissionError: [Errno 13] Permission denied: 'C:\Users\len\ANACON~1\envs\r-tensorflow\Lib\site-packages\numpy\core\multiarray.cp36-win_amd64.pyd' You are using pip version 9.0.1, however version 18.1 is available. You should consider upgrading via the 'python -m pip install --upgrade pip' command. Error: Error 2 occurred installing packages into conda environment r-tensorflow