[conda] Managing different package versions

[Jonathan Szlachta]

I'm interested in knowing how you handle installing different package versions. Do you use the idea of modules, different environments, or any other idea? What is the best approach in this case?

[Mumtaz Gohar]

You can just install the Anaconda latest version and install the related
packages through Anaconda prompt. When the installation of Anaconda package
manager is finished then start Anaconda prompt and write the following
command.
pip install package name then press enter.
Example
pip install keras
Note: Anaconda package manager is the best option to install any package
for python language.

Post by Jonathan Szlachta
I'm interested in knowing how you handle installing different package
versions. Do you use the idea of modules, different environments, or any
other idea? What is the best approach in this case?
--
You received this message because you are subscribed to the Google Groups
"conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.
To view this discussion on the web visit https://groups.google.com/a/
continuum.io/d/msgid/conda/886951fe-84ae-4833-adf2-
51b74fb0543d%40continuum.io.
For more options, visit https://groups.google.com/a/continuum.io/d/optout.

[Jonathan Rodrigues]

Cool. But let's say that I want to install numpy 1.11.3 and 1.12.1 alongside. Am I able to do that and choose which version I want?

[Paul Hobson]

Yes.

For that I would create new environments.

conda create --name=numpy112 python=3.6 numpy=1.12

and

conda create --name=numpy111 python=3.6 numpy=1.11

I don't install anything with pip except as a second-to-last resort (last
resort is building from source code).

On Mon, Jan 22, 2018 at 7:43 AM, Jonathan Rodrigues <

Post by Jonathan Rodrigues
Cool. But let's say that I want to install numpy 1.11.3 and 1.12.1
alongside. Am I able to do that and choose which version I want?
--
You received this message because you are subscribed to the Google Groups
"conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.
To view this discussion on the web visit https://groups.google.com/a/
continuum.io/d/msgid/conda/f5d4beed-f2e0-49c3-95cc-
d9a921617220%40continuum.io.
For more options, visit https://groups.google.com/a/continuum.io/d/optout.

[Dharhas Pothina]

Going one step further. I would get in the habit of using environment.yml
files to setup your environments. This way you know exactly what you
specified to create the environment you are using for a particular project
and can easily recreate it when necessary.

An example from one of my projects, that uses multiple conda channels and
also has pip dependencies:

https://github.com/ulmo-dev/ulmo/blob/master/py3_conda_environment.yml

Now I can create my env either by typing:

conda env create --file py3_conda_environment.yml (this creates a conda env
called py3-ulmo)

or

conda env create -n myenvname --file py3_conda_environment.yml . (if I want
to override my default env name)

Rules of thumb.

- always try to use conda packages over pip
- if a conda package is not available but the module is pure python, then
pip works fine.
- if package is not pure python then pip installing it may not work reliably

- dharhas

Post by Paul Hobson
Yes.
For that I would create new environments.
conda create --name=numpy112 python=3.6 numpy=1.12
and
conda create --name=numpy111 python=3.6 numpy=1.11
I don't install anything with pip except as a second-to-last resort (last
resort is building from source code).
On Mon, Jan 22, 2018 at 7:43 AM, Jonathan Rodrigues <

Post by Jonathan Rodrigues
Cool. But let's say that I want to install numpy 1.11.3 and 1.12.1
alongside. Am I able to do that and choose which version I want?
--
You received this message because you are subscribed to the Google Groups
"conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at https://groups.google.com/a/continuum.io/group/conda/
.
To view this discussion on the web visit
https://groups.google.com/a/continuum.io/d/msgid/conda/f5d4beed-f2e0-49c3-95cc-d9a921617220%40continuum.io
.
For more options, visit https://groups.google.com/a/continuum.io/d/optout
.

--
You received this message because you are subscribed to the Google Groups
"conda - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an
Visit this group at https://groups.google.com/a/continuum.io/group/conda/.
To view this discussion on the web visit
https://groups.google.com/a/continuum.io/d/msgid/conda/CADT3MED_oTbKXZapd9DkuHSc7JcB2HLKrMHowDg5HAfn5zj-GA%40mail.gmail.com
<https://groups.google.com/a/continuum.io/d/msgid/conda/CADT3MED_oTbKXZapd9DkuHSc7JcB2HLKrMHowDg5HAfn5zj-GA%40mail.gmail.com?utm_medium=email&utm_source=footer>
.
For more options, visit https://groups.google.com/a/continuum.io/d/optout.