jjhelmus / berryconda

Conda based Python distribution for the Raspberry Pi

This project is no longer active, no recipes will be updated and no packages will be added to the rpi channel.

Please use issue #83 to discuss the future of the project.

Berry Conda

Berryconda is a conda based Python distribution for the Raspberry Pi. With it, you can install and manage a scientific or Pydata stack on your Raspberry Pi using conda, a package and environment management system. All this can be done without compiling a single package!

Quick start

Berryconda is designed to work with raspbian jessie. Other Linux versions and distributions may or may not work. When using non-raspbian based distributions, bzip2 should be installed.

To install Berryconda, download the installer appropiate for your Raspberry Pi model. For Raspberry Pi 2 or 3 use the armv7l installers. For Raspberry Pi 1 or Zero use the armv6l installer.

Berryconda comes in two 'flavors', Berryconda2 and Berryconda3. The difference between these are the version of Python installed; Berryconda2 installs Python 2.7, and Berryconda3 installs Python 3.6. Choose the version you want installed.

armv7l installers (Raspberry Pi 2 or 3)

• Berryconda3-2.0.0-Linux-armv7l.sh
• Berryconda2-2.0.0-Linux-armv7l.sh

armv6l installers (Raspberry Pi 1 or Zero)

• Berryconda3-2.0.0-Linux-armv6l.sh
• Berryconda2-2.0.0-Linux-armv6l.sh

Once this file is downloaded on your Raspberry Pi, make the file executable using chmod and the execuate the installer. For example, to install Berryconda3 on a Raspberry Pi 3:

chmod +x Berryconda3-2.0.0-Linux-armv7l.sh
./Berryconda3-2.0.0-Linux-armv7l.sh

Follow the prompts to finish your install of Berryconda.

Berryconda should be installed in a user account, not by root or using sudo.

Once installed use the conda command to add packages from the rpi channel.

Details

Berryconda is created using constructor using the configuration files in the installer directory of this repository.

The packages in the rpi channel were created using conda-build using the recipes in the recipes directory of this repository.

Package building is done on two Raspberry Pi 3s, a Raspberry Pi 1, and a Raspberry Pi Zero.

Issues and package requests

Please report any problems with Berryconda or the packages in the rpi channel by submitting an issue. Also, use this link to request new packages.

Acknowlegements

The majority of the packages used in Berryconda were adapted from conda-forge recipes. A big thanks to everyone involved in the project!

Thanks to Continuum Analytics for hosting the rpi channel on the Anaconda Cloud and for creating the conda ecosystem.

berryconda's Issues

Including pytorch?

At the moment instally pytorch (cpu support, no cuda) can only be done installing from source which takes at least 8 hours! It would be great to have a binary!

bin/conda being removed

Thank you very much for Berryconda.

There is a problem in installing conda install -c rpi conda-build=2.1.10 in RPi B, having raspbian jessie 8 with Berryconda2-1.0.0-Linux-armv7l.sh. The error is as follows

ERROR conda.core.link:_execute_actions(335): An error occurred while installing package 'defaults::conda-build-2.1.10-py27_0'.
IOError(2, 'No such file or directory')
Attempting to roll back.
IOError(2, 'No such file or directory')

More than the error, further try for the conda command ends in -bash: /home/pi/berryconda2/bin/conda: No such file or directory, making to suspect the installation command completely removes the conda bin from the berryconda installation. To get back the lost bin/conda, a fresh installation of the berryconda is followed. Similar losing of bin/conda was observed for installing packages such as anaconda-client etc too.

OPencv3 with libgtk2 bindings

Do you tink it is possible to have this package with these bindings available?
I tried to compile opencv on my own to no avail.
Thanks for all your hard work!

Install ipyleaflet

Having ipyleaflet would be awesome giving its recent improvements.

More generally speaking, maybe it would be helpful to provide a section about how to contribute such additional packages as external contributions to this project? Thanks a lot, BTW!

conda run: qemu uncaught target signal 4 error

Hi!
I am trying to create armhf Dockerfile with automated build of the JupyterLab using this hack: https://resin.io/blog/building-arm-containers-on-any-x86-machine-even-dockerhub/

Here's the link to my repo: https://github.com/movalex/rpi-jupyter-conda/blob/base/Dockerfile

Unfortunately when Berryconda3-2.0.0-Linux-armv7l.sh is installed, I've got the error on this step:

conda config --system --add channels rpi

qemu: uncaught target signal 4 (Illegal instruction) - core dumped
Illegal instruction

Is it possible to fix this error?

ipython qtconsole

Hi,

I just successfully installed berryconda on the pi, followed by numpy, matplotlib and jupyter notebook. However, I can't find ipython qtconsole, is it because it is not in the default repo ? or another reason ?

thanks

how to source berryconda in debian

The installation instructions are missing a step, I said "yes" to add anaconda reference links in the bashrc file however this is not as easy and simple as ubuntu where I have anaconda working no problem. In debian the bashrc file is in a different location and the typical source command doesn't work in debian so I need some assistance on which command or where the bashrc file is located on the raspberry pi

Issue with Berryconda cryptography and OpenSSL

I'm trying to get paramiko working with the Berryian 2.0.0 distribution. Using cryptography 2.3.1, OpenSSL 1.0.2p, and paramiko 2.4.2 to open an SSH connection givens the error:

cryptography.exceptions.UnsupportedAlgorithm: Backend object does not support secp256r1.

I checked with the cryptography contributors, and they said that the OpenSSL backend should provide support for secp256r1. See pyca/cryptography#4691. I'm wondering if there's any known reason why this might be happening.

Versions of Python, cryptography, cffi, pip, and setuptools you're using

Python: 3.6.1
cryptography: 2.3.1
cffi: 1.11.5
pip: 18.1
setup tools: 36.0.1
openssl: 1.0.2p

Running on Raspberry Pi 3+ with Raspbian Stretch and the Berryconda 2.0.0 distribution.

Clear steps for reproducing your bug

Installed Berryconda; pip installed paramiko; ran a simple example:

import paramiko
t = paramiko.Transport(("127.0.0.1", 22))
t.connect(username="pi", password=<password>)

This gives error:

Unknown exception: Backend object does not support secp256r1.
Traceback (most recent call last):
  File "/home/producer/berryconda3/lib/python3.6/site-packages/paramiko/transport.py", line 2010, in run
    self._handler_table[ptype](self, m)
  File "/home/producer/berryconda3/lib/python3.6/site-packages/paramiko/transport.py", line 2126, in _negotiate_keys
    self.kex_engine.start_kex()
  File "/home/producer/berryconda3/lib/python3.6/site-packages/paramiko/kex_ecdh_nist.py", line 32, in start_kex
    self._generate_key_pair()
  File "/home/producer/berryconda3/lib/python3.6/site-packages/paramiko/kex_ecdh_nist.py", line 53, in _generate_key_pair
    self.P = ec.generate_private_key(self.curve, default_backend())
  File "/home/producer/berryconda3/lib/python3.6/site-packages/cryptography/hazmat/primitives/asymmetric/ec.py", line 287, in generate_private_key
    return backend.generate_elliptic_curve_private_key(curve)
  File "/home/producer/berryconda3/lib/python3.6/site-packages/cryptography/hazmat/backends/openssl/backend.py", line 1346, in generate_elliptic_curve_private_key
    _Reasons.UNSUPPORTED_ELLIPTIC_CURVE
cryptography.exceptions.UnsupportedAlgorithm: Backend object does not support secp256r1.

There is a a small error on numba's official web site

In this page ,the link to Berryconda is wrong to the https://https//github.com/jjhelmus/berryconda

Numba installing failed

Hello:

Thank you

I tried to install numba using berryconda but i obtained this error:

Loading channels: done

PackagesNotFoundError: The following packages are not available from current channels:

• numba

Current channels:

• https://conda.anaconda.org/rpi/linux-armv7l
• https://conda.anaconda.org/rpi/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.

Any suggestion?

Pip under berryconda behaves differently than in does under anaconda.

First. thanks very much for creating this package!

Next: How do I make pip play nice with berryconda? I have a project on a Pi3 that needs setproctitle. It's not available thru conda. On OS X using Anaconda I'm used to being able to install packages with pip and have all the paths come out correctly in the virtual environment.

The terminal session below shows some system info and my path in a Python 3.5 root environment. Then I switch to a Python 2.7 environment and show the path. Trying pip install setproctitle says it's already available under /usr/local/lib.

Then I start Python, check sys.path and then try to import setproctitle but get an ImportError -- which makes sense because sys.path only contains the pwd and the berryconda paths.

What's the recommended fix?

Thanks,
Mike

py27) pi@pi2:~ $ uname -a
Linux pi2 4.4.34-v7+ #930 SMP Wed Nov 23 15:20:41 GMT 2016 armv7l GNU/Linux

pi@pi2:~ $ echo $PATH
/home/pi/berryconda3/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games

pi@pi2:~ $ source activate py27
(py27) pi@pi2:~ $ echo $PATH
/home/pi/berryconda3/envs/py27/bin:/home/pi/berryconda3/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games
(py27) pi@pi2:~ $ which pip
/usr/local/bin/pip

(py27) pi@pi2:~ $ pip install setproctitle
Requirement already satisfied: setproctitle in /usr/local/lib/python2.7/dist-packages

(py27) pi@pi2:~ $ python
Python 2.7.13 | packaged by rpi | (default, Jan 15 2017, 19:52:30) 
[GCC 4.9.2] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.path
['', '/home/pi/berryconda3/envs/py27/lib/python27.zip', '/home/pi/berryconda3/envs/py27/lib/python2.7', '/home/pi/berryconda3/envs/py27/lib/python2.7/plat-linux2', '/home/pi/berryconda3/envs/py27/lib/python2.7/lib-tk', '/home/pi/berryconda3/envs/py27/lib/python2.7/lib-old', '/home/pi/berryconda3/envs/py27/lib/python2.7/lib-dynload', '/home/pi/berryconda3/envs/py27/lib/python2.7/site-packages']

>>> import setproctitle
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: No module named setproctitle
>>> 

opencv 3.3.0: python 3.5: ImportError: numpy.core.multiarray failed to import

I used berryconda to install opencv 3.3.0 into a python 3.5 virtual environment
but "import cv2" yielded this error:
ImportError: numpy.core.multiarray failed to import

missing seaborn 0.8.1

seaborn 0.8.1 -missing. Can this be added?

//Rajesh

Importing Numpy on Raspberry Pi Zero W throws an error

So I got a new Raspberry Pi Zero W and tried getting Berryconda to run on it. I encountered an error when trying to import Numpy although the install seemed to have run without issues.
Please find the error and outputs of conda info and conda list below:

pi@raspberrypi:~/berryconda3/pkgs $ which python
/home/pi/berryconda3/bin/python
pi@raspberrypi:~/berryconda3/pkgs $ python
Python 3.6.1 | packaged by rpi | (default, Apr 21 2017, 01:13:08) 
[GCC 4.9.2] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
Traceback (most recent call last):
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/core/__init__.py", line 16, in 
<module>
    from . import multiarray
ImportError: libgfortran.so.3: cannot open shared object file: No such file or directory

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/__init__.py", line 142, in <module>
    from . import add_newdocs
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/add_newdocs.py", line 13, in 
 <module>
    from numpy.lib import add_newdoc
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/lib/__init__.py", line 8, in <module>
     from .type_check import *
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/lib/type_check.py", line 11, in 
<module>
    import numpy.core.numeric as _nx
  File "/home/pi/berryconda3/lib/python3.6/site-packages/numpy/core/__init__.py", line 26, in 
<module>
    raise ImportError(msg)
ImportError: 
Importing the multiarray numpy extension module failed.  Most
likely you are trying to import a failed build of numpy.
If you're working with a numpy git repo, try `git clean -xdf` (removes all
files not under version control).  Otherwise reinstall numpy.

Original error was: libgfortran.so.3: cannot open shared object file: No such file or directory

Output of conda info:

pi@raspberrypi:~/berryconda3/pkgs $ conda info
Current conda install:

           platform : linux-armv6l
      conda version : 4.3.22
   conda is private : False
  conda-env version : 4.3.22
conda-build version : not installed
     python version : 3.6.1.final.0
   requests version : 2.18.1
   root environment : /home/pi/berryconda3  (writable)
default environment : /home/pi/berryconda3
   envs directories : /home/pi/berryconda3/envs
                      /home/pi/.conda/envs
      package cache : /home/pi/berryconda3/pkgs
                      /home/pi/.conda/pkgs
       channel URLs : https://conda.anaconda.org/rpi/linux-armv6l
                      https://conda.anaconda.org/rpi/noarch
        config file : /home/pi/berryconda3/.condarc
         netrc file : None
       offline mode : False
         user-agent : conda/4.3.22 requests/2.18.1 CPython/3.6.1 Linux/4.9.35+ debian/8.0 glibc/2.19    
            UID:GID : 1000:1000

Output of conda list:

pi@raspberrypi:~/berryconda3/pkgs $ conda list
# packages in environment at /home/pi/berryconda3:
#
asn1crypto                0.22.0                   py36_0  
beautifulsoup4            4.6.0                    py36_1  
ca-certificates           2017.4.17                     0  
certifi                   2017.4.17                py36_0  
cffi                      1.10.0                   py36_0  
chardet                   3.0.4                    py36_0  
conda                     4.3.22                   py36_0  
conda-env                 2.6.0                         0  
cryptography              1.9                      py36_0  
cython                    0.26                     py36_0  
idna                      2.5                      py36_0  
libffi                    3.2.1                         0  
libgfortran               3.0.0                         0  
ncurses                   5.9                           0  
numpy                     1.13.1                   py36_0  
openblas                  0.2.19                        0  
openssl                   1.0.2l                        0  
packaging                 16.8                     py36_0  
pip                       9.0.1                    py36_0  
pycosat                   0.6.1                    py36_0  
pycparser                 2.17                     py36_0  
pyopenssl                 16.2.0                   py36_0  
pyparsing                 2.2.0                    py36_0  
python                    3.6.1                         0  
python-dateutil           2.6.1                      py_0  
pytz                      2017.2                     py_0  
readline                  6.3                           0  
requests                  2.18.1                   py36_0  
ruamel_yaml               0.11.14                  py36_0  
setuptools                36.0.1                   py36_0  
six                       1.10.0                   py36_0  
sqlite                    3.15.0                        0  
tk                        8.6.6                         0  
urllib3                   1.21.1                   py36_0  
wheel                     0.29.0                   py36_0  
xz                        5.2.2                         0  
yaml                      0.1.6                         0  
zlib                      1.2.11                        0

ARMv6 also appears in Raspberry Pi2

Hi, you mention in your manual that Pi2 will need the AMRv7 version; however, I just realized that this might not be the case for all Pi2. I got an "illegal instruction" error when trying to use conda and by googling discovered that AMRv7 was the culprit. With AMRv6 everything runs smoothly (thanks btw!). I'd suggest adding "cat /proc/cpuinfo" to the manual to determine which version is needed.

ipywidgets

Using dask.distributed in the notebook requires the ipywidgets package to show progress bars.

opencv 3.4.2 release date

Do you know when version v3.4.2 for OpenCV is going to be made available? I'm saying this because I'm interested in using a new feature that came with this version called EAST text detector.

Here's an article of what I'm talking about:
https://www.pyimagesearch.com/2018/08/20/opencv-text-detection-east-text-detector/

Thank you!

BeagleBone Black compatibility (vfpv4)

I am using Conda on an ARM board quite similar to the Raspberry Pi, the BeagleBone Black, and it would be quite convenient to be able to use the Berryconda channel for this – which almost works. The problem is the following:

The board has an AM335x 1GHz Cortex-A8 CPU and the stock Debian (jessie) distribution is armv7l with hardware floating point, which should be compatible with Raspbian. However, while the CPU supports NEON, it doesn't support VFPv4, which is enabled on the Berryconda armv7l binaries. (Symptom: Python crashes with SIGILL due to an v4-only FMA instruction.)

Would it be possible to build the armv7l binaries without vfpv4? The performance difference should be unnoticable for typical berryconda use cases.

Alternatively, is there a way to force conda to use the armv6l binaries? These seem to work.

Tensorflow

Hi Jonathan,

After reading https://github.com/samjabrahams/tensorflow-on-raspberry-pi and seeing that you published Tensorflow in one of your other channels I am wondering how feasible it is to have Tensorflow in Berryconda. It seems a bit less straightforward than other packages.

Thanks.

JupyterLab update

As of right now, Jupyter Lab on the rpi channel is stuck in a pre-alpha version of a computational environment for Jupyter. Any chance an updated build could be done?

`illegal instruction` on conda on fresh install on raspberry pi zero w

Hi! Just did a fresh install of this on a Raspberry Pi Zero W and am seeing strange behavior.
I'm on Raspbarian Stretch Lite, 2018-06-27, and I just installed Berryconda. I've attached the terminal session where I discovered the issue.

Any invocation of conda dies with Illegal instruction after about a second.
Same with conda info conda install conda develop etc.
python works, and is 3.6, though I can't tell if it's the conda version

Welcome to Berryconda3 2.0.0

In order to continue the installation process, please review the license
agreement.
Please, press ENTER to continue
>>>
Copyright (c) 2016-2017 Jonathan J. Helmus
All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:

    * Redistributions of source code must retain the above copyright
       notice, this list of conditions and the following disclaimer.

    * Redistributions in binary form must reproduce the above
       copyright notice, this list of conditions and the following
       disclaimer in the documentation and/or other materials provided
       with the distribution.

    * Neither the name of the developers nor the names of any
       contributors may be used to endorse or promote products derived
       from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
        # (ISO/IEC-6429). (Lack of such support is extremely rare, and such


Do you approve the license terms? [yes|no]
[no] >>> y
Please answer 'yes' or 'no':
>>> yes

Berryconda3 will now be installed into this location:
/home/pi/berryconda3

  - Press ENTER to confirm the location
  - Press CTRL-C to abort the installation
  - Or specify a different location below

[/home/pi/berryconda3] >>>
PREFIX=/home/pi/berryconda3

installing: python-3.6.1-0 ...
^[[B^[[B^[[B^[[B^[[B^[[B^[[BPython 3.6.1
installing: asn1crypto-0.22.0-py36_0 ...
installing: ca-certificates-2017.4.17-0 ...
installing: certifi-2017.4.17-py36_0 ...
installing: cffi-1.10.0-py36_0 ...
installing: chardet-3.0.4-py36_0 ...
installing: conda-4.3.22-py36_0 ...
installing: conda-env-2.6.0-0 ...
installing: cryptography-1.9-py36_0 ...
installing: idna-2.5-py36_0 ...
installing: libffi-3.2.1-0 ...
installing: ncurses-5.9-0 ...
installing: openssl-1.0.2l-0 ...
installing: packaging-16.8-py36_0 ...
installing: pip-9.0.1-py36_0 ...
installing: pycosat-0.6.1-py36_0 ...
installing: pycparser-2.17-py36_0 ...
installing: pyopenssl-16.2.0-py36_0 ...
installing: pyparsing-2.2.0-py36_0 ...
installing: readline-6.3-0 ...
installing: requests-2.18.1-py36_0 ...
installing: ruamel_yaml-0.11.14-py36_0 ...
installing: setuptools-36.0.1-py36_0 ...
installing: six-1.10.0-py36_0 ...
installing: sqlite-3.15.0-0 ...
installing: tk-8.6.6-0 ...
installing: urllib3-1.21.1-py36_0 ...
installing: wheel-0.29.0-py36_0 ...
installing: xz-5.2.2-0 ...
installing: yaml-0.1.6-0 ...
installing: zlib-1.2.11-0 ...
installation finished.
Do you wish the installer to prepend the Berryconda3 install location
to PATH in your /home/pi/.bashrc ? [yes|no]
[no] >>>
You may wish to edit your .bashrc or prepend the Berryconda3 install location:

$ export PATH=/home/pi/berryconda3/bin:$PATH

Thank you for installing Berryconda3!
pi@demozero:~ $ less /.bash_rc
/.bash_rc: No such file or directory
pi@demozero:~ $ conda
-bash: conda: command not found
pi@demozero:~ $ vim .bashrc
pi@demozero:~ $ bash
pi@demozero:~ $ conda
Illegal instruction
pi@demozero:~ $ python --version
Python 3.6.1
pi@demozero:~ $ python
Python 3.6.1 | packaged by rpi | (default, Apr 20 2017, 19:35:19)
[GCC 4.9.2] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>>

pyserial

would it be possible to get the pyserial package?

opencv3 install

How we can install opencv3 ?

conda install -c menpo opencv3

gives

Solving environment: failed

PackagesNotFoundError: The following packages are not available from current channels:

• opencv3

Current channels:

• https://conda.anaconda.org/menpo/linux-armv7l
• https://conda.anaconda.org/menpo/noarch
• https://conda.anaconda.org/rpi/linux-armv7l
• https://conda.anaconda.org/rpi/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.

holoviews

Now that bokeh-charts has been deprecated in favor of holoviews (which is pretty awesome), it would be nice to have holoviews packages. I think to be useful also requires packages for param, paramnb, and parambokeh.

Package request: Py-ART

Hey @jjhelmus : how hard would it be it package Py-ART?
Just got my Pi today and think this could be a cool "Py-ART in a can" tool

including Pystan?

Any chance of seeing Pystan on Berryconda?

cannot install on aarch64 GNU/Linux

PREFIX=/home/simon/berryconda3
installing: python-3.6.1-0 ...
./Berryconda3-2.0.0-Linux-armv7l.sh: line 212: /home/simon/berryconda3/bin/python: No such file or directory
ERROR:
cannot execute native linux-armv7l binary, output from 'uname -a' is:
Linux localhost 4.9.65-perf+ #2 SMP PREEMPT Wed Jan 9 22:16:26 CST 2019 aarch64 aarch64 aarch64 GNU/Linux

OpenCV with imshow

In attempting to view images in openCV I run into this issue:

OpenCV Error: Unspecified error (The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Carbon support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script) in cvShowImage, file /home/jhelmus/conda/conda-bld/opencv_1512148862464/work/modules/highgui/src/window.cpp, line 605
Traceback (most recent call last):
  File "stereoCam.py", line 18, in <module>
    cv2.imshow('Camera', image)
cv2.error: /home/jhelmus/conda/conda-bld/opencv_1512148862464/work/modules/highgui/src/window.cpp:605: error: (-2) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Carbon support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function cvShowImage

So it looks like openCV wasn't compiled with UI support. Is this something that you could provide?

For completeness this is the code snippet that generates the error:

from picamera.array import PiRGBArray
from picamera import PiCamera
import time
import cv2

camera = PiCamera()
camera.resolution = (640, 480)
camera.framerate = 30
rawCapture = PiRGBArray(camera, size=camera.resolution)

# pause while the camera inits
time.sleep(0.1)

for frame in camera.capture_continuous(rawCapture, format='bgr', use_video_port=True):
    image = frame.array
    # display using opencv
    cv2.imshow('Camera', image)
    key = cv2.waitKey(1) & 0xFF
    # clear the stream for the next frame
    rawCapture.truncate(0)
    if key == ord('q'):
        break

install Tensorflow

I have installed tensorflow using conda and it did not show any error, but when I tried to check if the installation works properly it shows the following error:
no module named tensorflow

Import Pandas throws up an error

Importing Pandas on the raspberry pi, the following error crops up

`pi@raspberrypi:~/scripts $ python
 Python 3.6.1 | packaged by rpi | (default, Apr 20 2017, 19:35:19) 
 [GCC 4.9.2] on linux
 Type "help", "copyright", "credits" or "license" for more information.
 >>> import pandas
Traceback (most recent call last):
  File "/home/pi/berryconda3/lib/python3.6/site-packages/pandas/__init__.py", line 26, in <module>
from pandas._libs import (hashtable as _hashtable,
  File "/home/pi/berryconda3/lib/python3.6/site-packages/pandas/_libs/__init__.py", line 3, in <module>
    from .tslib import iNaT, NaT, Timestamp, Timedelta, OutOfBoundsDatetime
ImportError: /home/pi/berryconda3/lib/python3.6/site-packages/pandas/_libs/tslib.cpython-36m-arm-linux-gnueabihf.so: file too short
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/pi/berryconda3/lib/python3.6/site-packages/pandas/__init__.py", line 35, in <module>
    "the C extensions first.".format(module))
 ImportError: C extension: /home/pi/berryconda3/lib/python3.6/site-
 packages/pandas/_libs/tslib.cpython-36m-arm-linux-gnueabihf.so: file too short not built. If you want 
 to import pandas from the source directory, you may need to run 'python setup.py build_ext --inplace 
--force' to build the C extensions first.
>>>`

ERROR: size of Berryconda2-2.0.0-Linux-armv7l.sh should be 33565270 bytes

Hi,

I downloaded the file for Python 2 on a RaspberryPi 3B+ on my Mac and then used FTP to put it on the raspberry pi. I then made the .sh file executable using instructions provided but when I then try to execute it I get the following message:

ERROR: size of Berryconda2-2.0.0-Linux-armv7l.sh should be 33565270 bytes

Any idea what I a doing wrong?

JupyterHub

Hi,

I'm trying to install jupyterHub on raspberry, is there any chance to get it from this package?

Thank you very much!

xlrd package

Hello Jonathan,

I tried to import Excel data in Pandas (with : pd.read_excel) and got :
ImportError: No module named 'xlrd'

xlrd is not in your repo and also tried from conda but no success :
PackageNotFoundError: Package not found: '' Package missing in current linux-armv7l channels:

• xlrd

Thanks !!
(Thanks for you very useful repo on a Pi !! great time saver)

Package update request: Pandas

Hi there and thanks again for the great work.
I would like to kindly request an update for Pandas from 0.22.0 to the latest version (currently 0.23.1), if possible.

Edit: Bokeh 0.13.0 has just been released, too. This would be much appreciated as well.

Cheers

Update request: Bokeh 0.12.15

Please report any problems with Berryconda or the packages in the rpi channel by submitting an issue. Also, use this link to request new packages.

I would like to request Bokeh 0.12.15, since only 0.12.13 from December 2017 seems to be available on the rpi channel. I hope this is the right way to go about this. Or should I try to build Bokeh myself?

Many thanks!

Python 3.7

Python 3.7 was released last week. It would be awesome if berryconda3 could be updated to install it.

Jupyter in Berryconda - Native kernel (python3) is not available

There appears to be an issue running Jupyter notebook in Berryconda.

I'm using a Raspberry Pi 3 Model B, with Raspbian Jessie 8. Using Berryconda3-1.0.0-Linux-armv7l.sh.

Typing jupyter run or jupyter-run yields the following error:

pi@raspberrypi:~ $ jupyter-run
[RunApp] WARNING | Native kernel (python3) is not available
Traceback (most recent call last):
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/kernelspec.py", line 201, in get_kernel_spec
    resource_dir = d[kernel_name.lower()]
KeyError: 'python3'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/pi/berryconda3/bin/jupyter-run", line 6, in <module>
    sys.exit(jupyter_client.runapp.main())
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_core/application.py", line 267, in launch_instance
    return super(JupyterApp, cls).launch_instance(argv=argv, **kwargs)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/traitlets/config/application.py", line 657, in launch_instance
    app.initialize(argv)
  File "<decorator-gen-7>", line 2, in initialize
  File "/home/pi/berryconda3/lib/python3.5/site-packages/traitlets/config/application.py", line 87, in catch_config_error
    return method(app, *args, **kwargs)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/runapp.py", line 76, in initialize
    JupyterConsoleApp.initialize(self)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/consoleapp.py", line 334, in initialize
    self.init_kernel_manager()
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/consoleapp.py", line 288, in init_kernel_manager
    self.kernel_manager.start_kernel(**kwargs)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/manager.py", line 244, in start_kernel
    kernel_cmd = self.format_kernel_cmd(extra_arguments=extra_arguments)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/manager.py", line 175, in format_kernel_cmd
    cmd = self.kernel_spec.argv + extra_arguments
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/manager.py", line 87, in kernel_spec
    self._kernel_spec = self.kernel_spec_manager.get_kernel_spec(self.kernel_name)
  File "/home/pi/berryconda3/lib/python3.5/site-packages/jupyter_client/kernelspec.py", line 203, in get_kernel_spec
    raise NoSuchKernel(kernel_name)
jupyter_client.kernelspec.NoSuchKernel: No such kernel named python3

Package Request - pandas datareader (Berry Conda on OSMC)

I have a package build request for pandas-datareader so it will be available on the rpi channel (per the Anaconda docs here):

https://anaconda.org/anaconda/pandas-datareader

Also regarding compatibility you may wish to know that I successfully installed Berry Conda on OSMC, after installing bzip2, as described in this thread:

https://discourse.osmc.tv/t/can-i-install-berry-conda-python-package-manager-on-osmc/36630

Thank you for any attention you may be able to devote to this package request.

QT/PyQT/PyOpenGL

I'm wondering if it is possible to get Qt5, PyQt and PyOpenGL recipes ready.

There are some suggestions out there to compile Qt5 natively on raspberry PI 2&3. I would give it a try, but would need some guidance in which dependent packages need to come with berryconda and which are considered system-packages (coming with raspbian). The iteration cycle will be extremely slow, considering the approx. 2 days compilation time for Qt5 on an overclocked PI3 (https://wiki.qt.io/Native_Build_of_Qt5_on_a_Raspberry_Pi)

@jjhelmus In the end I would only need PyQt and PyOpenGL. So sharing your ideas would be very much welcome.

Update: Unfortunately Qt4 is needed also.

Installing IDE for berryconda like spyder

I've installed berryconda, and it was very successful in downloading some requisite software for spectrometer use. However, I'm looking for a script editor that I can easily modify and run code in that needs to be run through berryconda (for the spectrometer). I've tried installing spyder and python-editor, but had no luck (Packages missing in current channels). Any suggestions?
PS: I'm fairly new to all this, so please bear me! Maybe the solution is super easy.

Implement wxPython

Any chance you could implement wxPython?

I'm interested in creating some recipes for other Python packages to help add to the rpi channel, but I figured I'd throw in this request while I am learning to do so. Thanks!

Issue with jupyter notebook

I'm using Berryconda3-1.0.0-Linux-armv7l.sh in my Raspberry Pi 3 Model B.
Berryconda3 was installed successfully.
When I run :
jupyter
I get :
usage: jupyter [-h] [--version] [--config-dir] [--data-dir] [--runtime-dir] [--paths] [--json] [subcommand] jupyter: error: one of the arguments --version subcommand --config-dir --data-dir --runtime-dir --paths is required
And when I try to run :
jupyter notebook
I get this error :

Traceback (most recent call last):
  File "/home/pi/berryconda3/bin/jupyter-notebook", line 7, in <module>
    from notebook.notebookapp import main
  File "/home/pi/berryconda3/lib/python3.6/site-packages/notebook/notebookapp.py", line 46, in <module>
    from zmq.eventloop import ioloop
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/__init__.py", line 47, in <module>
    from zmq import backend
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/backend/__init__.py", line 40, in <module>
    reraise(*exc_info)
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/utils/sixcerpt.py", line 34, in reraise
    raise value
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/backend/__init__.py", line 27, in <module>
    _ns = select_backend(first)
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/backend/select.py", line 27, in select_backend
    mod = __import__(name, fromlist=public_api)
  File "/home/pi/berryconda3/lib/python3.6/site-packages/zmq/backend/cython/__init__.py", line 6, in <module>
    from . import (constants, error, message, context,
ImportError: /usr/lib/arm-linux-gnueabihf/libstdc++.so.6: version `CXXABI_1.3.9' not found (required by /home/pi/berryconda3/lib/python3.6/site-packages/zmq/backend/cython/../../../../../libzmq.so.5)

I installed jupyter notebook using conda, and I checked that it is installed and it is in (conda list)

How can I run jupyter notebook ?

cv2.waitKey(1) function not implemented

I ran Microsoft-ELL tutorial on Pi 0 W, and it seems the opencv package is not compiled properly.
My version of opencv is
opencv: 3.3.0-py36hcfffc66_0
It is installed by this command
conda install -c microsoft-ell/label/pi_zero opencv
And the packages installed are
Package plan for installation in environment /home/pi/berryconda3/envs/py36:
The following NEW packages will be INSTALLED:

bzip2:       1.0.6-1               lisaong
ffmpeg:      3.2.4-hf767ee3_0
giflib:      5.1.4-h3b1de47_0
hdf5:        1.8.17-0
jasper:      1.900.1-4             microsoft-ell
jpeg:        9b-0                  microsoft-ell
libgfortran: 3.0.0-1
libpng:      1.6.29-0              microsoft-ell
libtiff:     4.0.6-0
libwebp:     0.5.2-hcb3eeb6_7
numpy:       1.14.0-py36hebd9de5_0
openblas:    0.2.19-0
opencv:      3.3.0-py36hcfffc66_0
x264:        20131217-hc5d671e_0

Here are the errors.

OpenCV Error: Unspecified error (The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Carbon support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script) in cvWaitKey, file /home/jhelmus/conda/conda-bld/opencv_1512232341522/work/modules/highgui/src/window.cpp, line 676
Traceback (most recent call last):
File "tutorial.py", line 89, in
main()
File "tutorial.py", line 49, in main
while (cv2.waitKey(1) & 0xFF) == 0xFF:
cv2.error: /home/jhelmus/conda/conda-bld/opencv_1512232341522/work/modules/highgui/src/window.cpp:676: error: (-2) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Carbon support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function cvWaitKey

How about Biopython, NLTK and Gensim?

Hi Jonathan,

I'm keen on biomedical text/data mining and I'd like to suggest Biopython, Gensim and NLTK as candidates for the rpi-channel.

Best regards,

Daniel

Suggestion for new package

Hi jjhelmus,

Would it be possible to add Scikit-optimize (SKOPT) (https://github.com/scikit-optimize/scikit-optimize) to the rpi channel?
SKOPT is implementing a Bayesian Optimization routine for optimizing hyperparameters in e.g. machine learnings. Furthermore, it has proven succesfull in optimization of scientific experiments.

Many thanks for the berryconda work!
Best
/sqbl

Package request: OpenCV 3.2.0

Hi,

I want to use Berryconda to do my project and I need an OpenCV package. Would you mind to create an OpenCV package? Thank you.

`pip install -e` installs the source under the current working directory instead of under the environment directory

I'm used to the behaviour of pip in the system-provided python and virtualenv/virtualenvwrapper, where pip install -e a package from git would clone the repo into a subdirectory of the virtualenv. For instance, using virtualenvwrapper, I can do this:

~ $ mkvirtualenv foobarenv
~ $ pip install -e git+https://github.com/foobar/foobar.git
~ $ ls -ld /home/ksze/.virtualenvs/foobarenv/src/
total 8
drwxr-xr-x 8 ksze ksze 4096 Jul 31 10:38 foobar
-rw-r--r-- 1 ksze ksze 185 Jul 31 10:39 pip-delete-this-directory.txt

And foobar's git repo would be cloned into /home/ksze/.virtualenvs/foobar/src/foobar, regardless of where I run pip from.

However, using berryconda, pip install -e just clones to a subdirectory under the current working directory. This behaviour is quite annoying, not to mention being incompatible with what people expect when using a requirements file.

~ $ conda create -n foobar pip
~ $ source activate foobar
~ $ pip install -e git+https://github.com/foobar/foobar.git
~ $ ls -ld ./src/
total 8
drwxr-xr-x 8 ksze ksze 4096 Jul 31 10:41 foobar
-rw-r--r-- 1 ksze ksze 185 Jul 31 10:40 pip-delete-this-directory.txt

I'm not sure if this is an upstream conda issue or specific to berryconda, as I have never used the "vanilla" conda.

openssl error message after running 'conda update --all'

@:~ $ openssl
WARNING: can't open config file: /home/jhelmus/bc3_v7/conda-bld/openssl_1547610726478/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_place/ssl/openssl.cnf

Just curious how this differs from Miniconda

What would be reasons to choose the one or the other?

Support for arm64

Hi, it would be interesting to add support to RPI3b+ (with non-raspbian OS as ubuntu or suse) using the arm64 architecture.

Regards