I'm using python 3.6.3 on a windows 10 machine. I installed pydot and graphviz using pip install via:
py -m pip install pydot
py -m pip install graphviz
I also went to the graphviz website and downloaded and installed the windows version here: http://www.graphviz.org/Download_windows.php which default installed to program files(x86). But when I go to plot my model in keras, I still get the error saying I have to install pydot and graphviz and that the import failed. I can do
import pydot
import graphviz
on my python console just fine, they throw no errors. What else should I do to be able to graph my neural net?
The path(s) to the installed GraphViz executables (dot, neato, etc.) need to be in the PATH environment variable, in order for pydot to find them. pydot used to search for those executables in earlier versions, but not any more.
Also:
pydot is a Python package.
GraphViz is a collection of tools written in C for computing graph layouts
graphviz is a Python package entirely unrelated to pydot. These two Python packages do not interact in any way with each other. Installing one of them should suffice (together with GraphViz).
See also:
https://github.com/erocarrera/pydot/issues/126
https://github.com/erocarrera/pydot/commit/812e3c40dac1e0225391481073c64da5bafba93e#commitcomment-18236709
https://github.com/erocarrera/pydot/commit/812e3c40dac1e0225391481073c64da5bafba93e#commitcomment-18222580
and links from there.
Complementing #Ioannis answer, you have to install GraphViz executables via conda (conda install GraphViz).
For my case, after installing GraphViz I tried with the latest pydot (pip install pydot) and the error was resolved.
Just to complete the #dataLeo 's solution, Python 3 users can use pydotplus package instead of pydot-ng package. To summarize:
install pydot+graphviz and pydotplus by commands "conda install pydot" and "conda install -c conda-forge pydotplus".
Go to the vis_utils.py file and change line 11 from import pydot to
import pydotplus as pydot.
PS: You can locate the vis_utils.py file by checking help for plot_model command in ipython console, i.e. after from keras.utils import plot_model, type ??plot_model in ipython console.
Tested on Windows 10-64 bit with Anaconda python-3.6
I solved this problem by installing the packages with :
conda install graphviz
conda install pydot
conda install pydotplus
Related
I have pygame in pip environment and used it without problem.
But now I switched to conda environment (called tf-gpu) to have TF features also.
As I figured out the most used channel for conda version of pygame is cogsci.
But conda install -c cogsci pygame in terminal causes an error:
It seems there is a version incompatibility.
I then downloaded the original tar.bz2 file from conda, and tried conda install --offline C:\...(file path)
It downloaded and recognized by interpreter(python 3.8.8, conda('tf_gp' env)).
Then Pyhton imported pygame but error of there is no module "from python.base import *" proceeded it.(as it is in the __init__ file of package)
Are these problem due to old version of pygame? I tried other channels in anaconda but unsuccessful(also they were older).
And also i can not downgrade python because of many packages' prerequisites.
I also tried installing via pip(in that conda promp), but it said that the requirment has been satisfied(as i previoisly used pygame in pip env(python 3.8.2).
Is there any new conda version of pygame?
Should i remove previous pygame in pip and try pip install again in conda env?
Or is it possible to modify main package source code(old file)?
This issue usually comes in when you are running many versions of python and pygame together which causes some incompatibility. Another problem which may occur is that anaconda environment stop recognising the pygame as it can’t locate it.
My suggestion is that uninstall all the older versions of python and pygame and install the latest ones.
Or else you may shift to Visual studio code for better experience and lesser problems.
Thanks,
The problem was solved by uninstalling other python version that I had(3.6). and using pip version of pygame(which did not work when two python versions (3.6, 3.8) were on my system.)
I'm a little confused with package install in anaconda environment.
I can install my python packagse in the following ways.
open anaconda prompt and do : conda install tensorflow
launch an jupyter notebook from anaconda prompt, choose the default python kernel, and do:
!pip install tensorflow
Can someone tell me what is the difference between these two cases? Where are the python packages installed? What happens when I pip install python packages in default kernel in jupyter notebook?
I was able to install tensorflow in jupyter notebook in default kernel (python3), but trying to import tensorflow give me module note found error. Does anyone know why that happened? What is happening under the hood? Very much appreciate any help to clarification.
pip is the default package manager that ships with python. Conda is also a package manager, but it is third party. Conda was made especially for data science libraries. Libraries installed with conda usually give much better performance than pip. In pip, the packages are stored in python/scripts and conda stores them at /anaconda/pkgs/. As for the module not found error. I would need more information about it, but you can check out this video. I learnt how to install TensorFlow here and would highly recommend it.
after following the instruction in http://docs.h2o.ai/h2o/latest-stable/h2o-docs/downloading.html#install-in-python, I was able to install h2o v 3.16.0.2. I was also able to use command line instructions mentioned in the above website and test that it is working.
python
import h2o
h2o.init()
h2o.demo("glm")
However, when I launch anaconda spyder, I am not able to import h2o. How do I link the h2o I have installed and bring it into Spyder python?
Update:
I have already tried {conda install -c anaconda h2o } which is mentioned on Anaconda.org, but that installs older ver 3.10 of h2o and that did not work either.
thanks for your help.
I believe the problem is due to your python environment. When you install Anaconda, you need to use anaconda pip, to ensure that the installed packages are available in conda.
The short answer is you can install the lastest stable version of H2O (3.16.0.2 as of today) using conda via h2oai channel instead of anaconda channel (h2o maintains it's own channel):
conda install -c h2oai h2o
this should solve your issue.
But more generally, the packages will appear in conda if you use anaconda pip. You can check which pip is being used by doing
which pip
and making sure that the path to the pip is within your anaconda distribution; something like /home/<userdir>/anaconda/bin/pip instead of /usr/bin/pip
Same is true also for Python. Try checking if starting Python in terminal points to anaconda Python by doing which python. If that's not the case, than you would need to add the conda installation of Python to your PATH variable. Please refer to conda docs for instructions https://conda.io/docs/user-guide/install/index.html
It would have been helpful if you had included information regarding your operating system in the question.
I want to install the python-igraph package, but I am currently using python 3.6.1 and I don't find any installer for this new version of python. Do you know how can I install python-igraph for this version?
I have tried to install python-igraph for older versions from anaconda cloud but a version problem occurs.
!conda install -c vtraag python-igraph
Fetching package metadata .............
Solving package specifications: .
UnsatisfiableError: The following specifications were found to be in conflict:
- python 3.6*
- python-igraph -> python 3.5*
Use "conda info <package>" to see the dependencies for each package.`
I know that python-igraph exists for older python versions, but I don't want to overwrite the current python 3.6 version, so I think I need a new environment to install the older version and be able to use both.
I have already tried to create a new environment for python 3.5:
!conda create -n py35 python=3.5 ipykernel
But after 30min I didn't obtain any response...
I have also followed the same procedure to install a package for python 2.7 (from marufr contributor), but I get the same problem creating an environment for python 2.7.
Note: I am working from Jupyter notebook on Windows (win-64) and using Anaconda.
Update:
Solution found here:
https://medium.com/towards-data-science/environment-management-with-conda-python-2-3-b9961a8a5097
Since there is no python-igraph for python 3.6 yet, I have to use an older version.
First, from Anaconda Prompt, install the nb_conda_kernels package before creating the new environment:
conda install nb_conda_kernels
Then, create the environment where I will work with the older python version:
conda create -n py35 python=3.5 ipykernel
Finally, install the package through the wheel (found here: http://www.lfd.uci.edu/~gohlke/pythonlibs/#python-igraph)
pip install python_igraph‑0.7.1.post6‑cp35‑none‑win_amd64.whl (You have to this command in the same folder that you donwloaded the wheel!)
There is a Windows installer for igraph‘s Python interface on the Python Package Index. Download the one that is suitable for your Python version (currently there are binary packages for Python 2.6, Python 2.7 and Python 3.2, though it might change in the future). To test the installed package, launch your favourite Python IDE and type the following:
import igraph.test
igraph.test.run_tests()
The above commands run the bundled test cases to ensure that everything is fine with your igraph installation.
Now, you can download 'python_geohash‑0.8.5‑cp36‑cp36m‑win_amd64.whl'.
Use the following command to install:
pip install python_geohash‑0.8.5‑cp36‑cp36m‑win_amd64.whl
Then you can use python-igraph in your Python3.6
I tried outputting one of my DiGraphs as follows:
A=nx.drawing.nx_agraph.to_agraph(Graph)
nx.drawing.nx_agraph.write_dot(dot_filename)
However I got :
ImportError: ('requires pygraphviz ', 'http://pygraphviz.github.io/')
Anaconda command windows says
graphviz 2.38.0 0 is already installed..
Can anybody help who managed to use Graphviz through Anaconda?
You can try this:
conda install -c pdrops pygraphviz=1.2
Check conda pygraphviz package availability for your system. Previous answer seems to be a package available just for linux & osx 64. Give a try to
conda install -c marufr pygraphviz=1.3.1
this version shows in the details to be compatible with win32 & 64 as well as with linux and osx. You can find more packages from any user typing
conda install -c binstar binstar
Afterthat, for instance, to find all the pygraphviz packages in public channels code
binstar search -t conda pygraphviz
Finally find your platform compatible and install from command line as usual.