I am trying to plot a simple graph using pyplot, e.g.:
import matplotlib.pyplot as plt
plt.plot([1,2,3],[5,7,4])
plt.show()
but the figure does not appear and I get the following message:
UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure.
I saw in several places that one had to change the configuration of matplotlib using the following:
import matplotlib
matplotlib.use('TkAgg')
import matplotlib.pyplot as plt
I did this, but then got an error message because it cannot find a module:
ModuleNotFoundError: No module named 'tkinter'
Then, I tried to install "tkinter" using pip install tkinter (inside the virtual environment), but it does not find it:
Collecting tkinter
Could not find a version that satisfies the requirement tkinter (from versions: )
No matching distribution found for tkinter
I should also mention that I am running all this on Pycharm Community Edition IDE using a virtual environment, and that my operating system is Linux/Ubuntu 18.04.
I would like to know how I can solve this problem in order to be able to display the graph.
Solution 1: is to install the GUI backend tk
I found a solution to my problem (thanks to the help of ImportanceOfBeingErnest).
All I had to do was to install tkinter through the Linux bash terminal using the following command:
sudo apt-get install python3-tk
instead of installing it with pip or directly in the virtual environment in Pycharm.
Solution 2: install any of the matplotlib supported GUI backends
solution 1 works fine because you get a GUI backend... in this case the TkAgg
however you can also fix the issue by installing any of the matplolib GUI backends like Qt5Agg, GTKAgg, Qt4Agg, etc
for example pip install pyqt5 will fix the issue also
NOTE:
usually this error appears when you pip install matplotlib and you are trying to display a plot in a GUI window and you do not have a python module for GUI display.
The authors of matplotlib made the pypi software deps not depend on any GUI backend because some people need matplotlib without any GUI backend.
In my case, the error message was implying that I was working in a headless console. So plt.show() could not work. What worked was calling plt.savefig:
import matplotlib.pyplot as plt
plt.plot([1, 2, 3], [5, 7, 4])
plt.savefig("mygraph.png")
I found the answer on a github repository.
If you use Arch Linux (distributions like Manjaro or Antegros) simply type:
sudo pacman -S tk
And all will work perfectly!
Simple install
pip3 install PyQt5==5.9.2
It works for me.
Try import tkinter because pycharm already installed tkinter for you, I looked Install tkinter for Python
You can maybe try:
import tkinter
import matplotlib
import matplotlib.pyplot as plt
matplotlib.use('TkAgg')
plt.plot([1,2,3],[5,7,4])
plt.show()
as a tkinter-installing way
I've tried your way, it seems no error to run at my computer, it successfully shows the figure. maybe because pycharm have tkinter as a system package, so u don't need to install it. But if u can't find tkinter inside, you can go to Tkdocs to see the way of installing tkinter, as it mentions, tkinter is a core package for python.
I added %matplotlib inline
and my plot showed up in Jupyter Notebook.
The answer has been given a few times but it is not obvious, one needs to install graphics, this works.
pip3 install PyQt5
I too had this issue in PyCharm. This issue is because you don't have tkinter module in your machine.
To install follow the steps given below (select your appropriate os)
For ubuntu users
sudo apt-get install python-tk
or
sudo apt-get install python3-tk
For Centos users
sudo yum install python-tkinter
or
sudo yum install python3-tkinter
for Arch Users
sudo pacman -S tk
or
sudo pamac install tk
For Windows, use pip to install tk
After installing tkinter restart your Pycharm and run your code, it will work
This worked with R reticulate. Found it here.
1: matplotlib.use( 'tkagg' )
or
2: matplotlib$use( 'tkagg' )
For example:
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import style
import matplotlib
matplotlib.use( 'tkagg' )
style.use("ggplot")
from sklearn import svm
x = [1, 5, 1.5, 8, 1, 9]
y = [2, 8, 1.8, 8, 0.6, 11]
plt.scatter(x,y)
plt.show()
If using Jupyter notebook try the following:
%matplotlib inline
This should render the plot even if not specifying the
plt.show()
command.
None of these answers worked for me using Pycharm Professional edition 2021.3
Regular matplotlib graphs did work on the scientific view, but it did not allow me to add images to the plots.
What did work for me is adding this line before I try plotting anything:
plt.switch_backend('TkAgg')
issue = “UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure.”
And this worked for me
import matplotlib
import matplotlib.pyplot as plt
matplotlib.use('Qt5Agg')
For Windows 10, if using pip install tk does not work for you, try:
Download and run official python installer for windows. Even if you
already have it downloaded, run it again.
When (re)installing python, make sure you chose "advanced" options, and
set the checkbox "tcl/tk and IDLE" to true.
If you already had python installed, select the "Modify" option, and
make sure that checkbox is selected.
Source of my fix:
https://stackoverflow.com/a/59970646/2506354
I have solved it by putting matplotlib.use('TkAgg') after all import statements.
I use python 3.8.5 VSCODE and anaconda.
No other tricks worked.
I installed python3-tk , on Ubuntu 20.04 and using WSL2
import numpy as np
import matplotlib.pyplot as plt
import matplotlib
matplotlib.use( 'tkagg')
and then I installed GWSL from the Windows Store which seems to solve problem of WSL2 rendering out of the box
This will solve the issue. It works well in jupyter.
%matplotlib inline
The comment by #xicocaio should be highlighted.
tkinter is python version-specific in the sense that sudo apt-get install python3-tk will install tkinter exclusively for your default version of python. Suppose you have different python versions within various virtual environments, you will have to install tkinter for the desired python version used in that virtual environment. For example, sudo apt-get install python3.7-tk. Not doing this will still lead to No module named ' tkinter' errors, even after installing it for the global python version.
On Mac OS, I made it work with:
import matplotlib
matplotlib.use('MacOSX')
Ubuntu 20.04 command line setup. I install the following to make Matplotlib stop throwing the error UserWarning: Matplotlib is currently using agg, which is a non-GUI backend, so cannot show the figure.
I installed python-tk through the steps:
apt-get update
apt-get install python3.8-tk
Just in case if this helps anybody.
Python version: 3.7.7
platform: Ubuntu 18.04.4 LTS
This came with default python version 3.6.9, however I had installed my own 3.7.7 version python on it (installed building it from source)
tkinter was not working even when the help('module') shows tkinter in the list.
The following steps worked for me:
sudo apt-get install tk-dev.
rebuild the python:
1. Navigate to your python folder and run the checks:
cd Python-3.7.7
sudo ./configure --enable-optimizations
Build using make command:
sudo make -j 8 --- here 8 are the number of processors, check yours using nproc command.
Installing using:
sudo make altinstall
Don't use sudo make install, it will overwrite default 3.6.9 version, which might be messy later.
Check tkinter now
python3.7 -m tkinter
A windows box will pop up, your tkinter is ready now.
After upgrading lots of packages (Spyder 3 to 4, Keras and Tensorflow and lots of their dependencies), I had the same problem today! I cannot figure out what happened; but the (conda-based) virtual environment that kept using Spyder 3 did not have the problem. Although installing tkinter or changing the backend, via matplotlib.use('TkAgg) as shown above, or this nice post on how to change the backend, might well resolve the problem, I don't see these as rigid solutions. For me, uninstalling matplotlib and reinstalling it was magic and the problem was solved.
pip uninstall matplotlib
... then, install
pip install matplotlib
From all the above, this could be a package management problem, and BTW, I use both conda and pip, whenever feasible.
You can change the matplotlib using backend using the from agg to Tkinter TKAgg using command
matplotlib.use('TKAgg',warn=False, force=True)
Works if you use some third party code in your project. It probably contains the following line
matplotlib.use('Agg')
Search for it and comment it out.
If you have no clue about what it is you are probably not using this part of the code.
Solutions about using another backend GUI may be cleaner, so choose your fighter.
The solution that worked for me:
Install tkinter
import tkinter into the module
make sure that matplotlib uses (TkAgg) instead of (Agg)
matplotlib.use('TkAgg')
execute the following command before plotting
%matplotlib inline
Try:
%matplotlib inline
I had the same problem and it worked for me. I tested it on my Jupyter notebooks and visual studio code, so you should have no problems.
On WSL with X server
Make sure that your X server work. Matplotlib indicate this error if he can't connect to the X display.
Windows Firewall configuration
Pay attention to the windows firewall ! I changed from WSL Debian to Ubuntu and didn't remember about the firewall rule.
I use this post to configure the windows firewall rule to make the X server work. This method avoid too permisive rule that able anyone to use your X server.
It said :
If you already had installed an X11 server, Windows may have created firewall rules that will mess with the above configuration. Search for them and delete them in "Windows Defender Firewall with Advanced Security."
You will now need to configure Windows Firewall to permit connections from WSL2 to the X11 display server. You will install the display server in the next step. We do this step first to avoid Windows Firewall from auto-creating an insecure firewall rule when you run the X11 display server. Many guides on X11 forwarding and WSL2 make this firewall rule too permissive, allowing connections from any computer to your computer. This means someone could theoretically, if they are on your same network, start sending graphical display information to your computer.
To avoid this, we will make Windows Firewall only accept internet traffic from the WSL2 instance.
To set this up, you can copy the below to a script and run it from within WSL2:
#!/bin/sh
LINUX_IP=$(ip addr | awk '/inet / && !/127.0.0.1/ {split($2,a,"/"); print a[1]}')
WINDOWS_IP=$(ip route | awk '/^default/ {print $3}')
# Elevate to administrator status then run netsh to add firewall rule
powershell.exe -Command "Start-Process netsh.exe -ArgumentList \"advfirewall firewall add rule name=X11-Forwarding dir=in action=allow program=%ProgramFiles%\VcXsrv\vcxsrv.exe localip=$WINDOWS_IP remoteip=$LINUX_IP localport=6000 protocol=tcp\" -Verb RunAs"
Manual method :
Alternatively, you can manually add the rule through a GUI by doing the following:
Open "Windows Defender Firewall with Advanced Security"
Click add new rule brings up the New Rule Wizard (next to navigate between each section):
Rule type: Custom
Program: "This program path:" %ProgramFiles%\VcXsrv\vcxsrv.exe
Protocol and ports
Protocol type: TCP
Local port: 6000
Remote port: any
Scope
Local IP address: Obtain the IP address to put in by running the below command in WSL2
ip route | awk '/^default/ {print $3}'
remote IP addresses
Obtain IP address to enter by running the below in WSL2
ip addr | awk '/inet / && !/127.0.0.1/ {split($2,a,"/"); print a[1]}'
Action: "Allow the connection
Profile: Selection Domain, Private, and Public
Name: "X11 forwarding"
Linux Mint 19. Helped for me:
sudo apt install tk-dev
P.S. Recompile python interpreter after package install.
When I ran into this error on Spyder, I changed from running my code line by line to highlighting my block of plotting code and running that all at once. Voila, the image appeared.
If you install python versions using pyenv on Debian-based systems, be sure to run sudo apt install tk-dev before pyenv install. If it's already installed, remove it with pyenv uninstall and install it again after install tk-dev. Therefore, there is no need to set any env variables when running pyenv install.
Alpaca backtrader plot issue: I ran into this import issue and found this article, so I applied the code, but same issue not resolved. any one can help please?
My installed matplotlib version is 3.3.1
backtrader 1.9.76.123
python 3.8.5
the entire code posted below:
from matplotlib.dates
import (HOURS_PER_DAY, MIN_PER_HOUR, SEC_PER_MIN,MONTHS_PER_YEAR,
DAYS_PER_WEEK,SEC_PER_HOUR, SEC_PER_DAY,num2date, rrulewrapper,
YearLocator,MicrosecondLocator)
import alpaca_backtrader_api
import backtrader as bt
from datetime import datetime
#import matplotlib
ALPACA_API_KEY = "XXXXX"
ALPACA_SECRET_KEY = "XXXX"
ALPACA_PAPER = True
class SmaCross(bt.SignalStrategy):
def init(self):
sma1, sma2 = bt.ind.SMA(period=10), bt.ind.SMA(period=30)
crossover = bt.ind.CrossOver(sma1, sma2)
self.signal_add(bt.SIGNAL_LONG, crossover)
cerebro = bt.Cerebro()
cerebro.addstrategy(SmaCross)
store = alpaca_backtrader_api.AlpacaStore( key_id=ALPACA_API_KEY,secret_key=ALPACA_SECRET_KEY,paper=ALPACA_PAPER)
if not ALPACA_PAPER:
broker = store.getbroker() # or just alpaca_backtrader_api.AlpacaBroker()
cerebro.setbroker(broker)
DataFactory = store.getdata # or use alpaca_backtrader_api.AlpacaData
data0 = DataFactory(dataname='AAPL', historical=True, fromdate=datetime(2015, 1, 1), timeframe=bt.TimeFrame.Days)
cerebro.adddata(data0)
print('Starting Portfolio Value: %.2f' % cerebro.broker.getvalue())
cerebro.run()
print('Final Portfolio Value: %.2f' % cerebro.broker.getvalue())
cerebro.plot()
Downgrade to matplotlib 3.2.2 until the bug in backtrader is fixed.
Here is the fix pull request: https://github.com/mementum/backtrader/pull/418.
pip uninstall matplotlib # or conda
pip install matplotlib==3.2.2
I suffered the same problem like you did, your link provided has the perfect solution. just get rid of warnings from locator.py
https://community.backtrader.com/topic/981/importerror-cannot-import-name-min_per_hour-when-trying-to-plot/8
I couldn't install matplotlib==3.2.2 nor the patch without uninstalling backtrader first.
So, this worked for me in the end:
Uninstall backtrader:
pip uninstall backtrader
Install the patch provided in the above solution:
pip install git+https://github.com/mementum/backtrader.git#0fa63ef4a35dc53cc7320813f8b15480c8f85517#egg=backtrader
If necessary, install matplotlib again:
pip install matplotlib
As pointed out above, the issue is addressed in this pull request and the patch is the latest commit to master, but there hasn't been a release since 2019-05.
You can install the patched version like so:
pip install git+https://github.com/mementum/backtrader.git#0fa63ef4a35dc53cc7320813f8b15480c8f85517#egg=backtrader
You could alternatively specify the required commit in requirements.txt like so:
-e git+https://github.com/mementum/backtrader.git#0fa63ef4a35dc53cc7320813f8b15480c8f85517#egg=backtrader
…then pip install -r requirements.txt
After installing with either method, you can confirm the versions installed with pip freeze:
...
backtrader==1.9.76.123
...
How to install from git
Mac Big Sur
for me it only worked if:
Downgrade python3.9 to python 3.8
then I downgraded matplotlib==3.2.2
For both python 3.8.x and 3.9.x, I solved the problem by using specific version of matplotlib==3.2.2
pip install matplotlib==3.2.2
By default, I used matplotlib==3.4.x version and the problem occured.
All of the above answers are fine. The problem is not with Matplotlib though. The Backtrader library hasn't kept up with the Matplotlib updates. You can do the off-label Backtrader update suggested by Joel Brigate above...or you can make a simple mod to locator.py file (backtrader.plot):
Just change:
from matplotlib.dates import (HOURS_PER_DAY, MIN_PER_HOUR,
SEC_PER_MIN, MONTHS_PER_YEAR, DAYS_PER_WEEK, SEC_PER_HOUR,
SEC_PER_DAY, num2date, rrulewrapper, YearLocator,
MicrosecondLocator, warnings)
to:
from matplotlib import warnings
from matplotlib.dates import (HOURS_PER_DAY, MIN_PER_HOUR, SEC_PER_MIN,
MONTHS_PER_YEAR, DAYS_PER_WEEK, SEC_PER_HOUR,
SEC_PER_DAY, num2date, rrulewrapper,
YearLocator, MicrosecondLocator)
You'll note that the warnings import now comes directly out of matplotlib rather than matplotlib.dates. This is the offending issue within locator.py.
Here is my solution:
python -m pip uninstall matplotlib
python -m pip uninstall backtrader
python -m pip install backtrader
python -m pip install matplotlib==3.2.2
Enjoy!
I could not install matplotlib==3.2.2 with python 3.9 .
Here is how did I fix this issue:
$ pip uninstall backtrader
$ pip install git+https://github.com/mementum/backtrader.git#0fa63ef4a35dc53cc7320813f8b15480c8f85517#egg=backtrader
Reference:
Github: Fix ImportError from matplotlib.dates
#418
Mac Big Sur. I did the same: python 3.8.5, uninstall matplotlib, install matplotlib==3.2.2
I'm new at this so I first tried the easy way, through anaconda.org, but could not find version 3.2.2. Then tried it from the Jupyter notebook with conda install... didn't work. I finally did it straight through terminal, which worked fine.
#laffuste solution of downgrading to version 3.2.2 of matplotlib solved the issue for me. PR to fix the issue is still open, you can also follow this forum for more info on the problem:
Jupyter Notebook kernel dies when importing tensorflow 1.5.0
I have read a lot of posts relating to this but they have all had higher version numbers of tensorflow and have solved it by downgrading to 1.5.0. I also had higher version number and followed the advice to downgrade but I still have the problem.
Does anyone know what to try next?
pip install h5py==2.8.0
worked for me
When trying using the command prompt I got an error message not related to the tensorflow issue (I think);
"Warning! HDF5 library version mismatched error"
The key information from that message body was "Headers are 1.10.1, library is 1.10.2" so I downgraded hdf5 library by "conda install -c anaconda hdf5=1.10.1" and now the error message is gone and the kernel does not die when importing tensorflow.
I got similar problems, any tensorflow or tensorflow related packages (e.g. keras) made my kernel to die when loading, from any interface (jupyter, spyder, console....)
For those having this kind of problems, try running python from the console with verbose mode (python -v) then import tensorflow and look for errors.
I spot errors related to h5py, similar to the reply of #DBSE. I just upgraded the h5py package then everything was solved !
If you are using a conda environment, then the easiest method for fixing this issue is to just create a new environment and install tensorflow with just a single command. I had the same issue, I have tried a lot on most of the version of python and tensorflow. But at the last I have successfully configured it with just a single steps.
Run this command for installing GPU version
conda create --name tf_gpu tensorflow-gpu
The above line of code will automatically install that version of python and tf which is comaptible with your GPU or CPU.
For CPU, Run this command
conda create --name tf_env tensorflow
Both of these command work 100 % with my system for GPU and CPU access and will download the latest version which are compatible with system. It will resolved/fixed "Illegal Instruction (code dumps)" error.
pip install h5py==3.1.0
This is the most updated version which worked for me.
Try using import numpy before Keras and Tensorflow.