search

cdmbi / PCM

PCM workflow
GNU General Public License v2.0
1 stars 4 forks source link

A script that sets up the environment is needed. #2

Open likit opened 9 years ago

likit commented 9 years ago

Installing conda-rdkit and create a conda environment for the pipeline is quite difficult for users. To facilitate the process, we should have a Python script that does all the work. I'll be working on that.

chaninn commented 9 years ago

Thanks for the heads up, that would be nice to have.

On Fri, Mar 27, 2015 at 6:11 PM, Likit Preeyanon notifications@github.com wrote:

Installing conda-rdkit and create a conda environment for the pipeline is quite difficult for users. To facilitate the process, we should have a Python script that does all the work. I'll be working on that.

— Reply to this email directly or view it on GitHub https://github.com/cdmbi/PCM/issues/2.

Associate Professor Chanin Nantasenamat, Ph.D. Head, Center of Data Mining and Biomedical Informatics Faculty of Medical Technology, Mahidol University 999 Phutthamonthon 4 Road, Salaya, Nakhon Pathom 73170 Tel. 02 441 4371 ext 2715 Fax 02 441 4380

likit commented 9 years ago

I uploaded install.py, which can download and build rdkit. However, upon further consideration, I'm now thinking of using conda to build a conda package and upload it to binstar instead. Likewise, I will use setuptools to make a Python package and upload it to pypi. This way users can use either conda or pip to install our package. Conda and setuptools can install specific version of all dependencies, so we won't have a problem when running our program on different machines/environments.

Here is a tutorial for conda build, if you're interested. http://conda.pydata.org/docs/build_tutorials.html

likit commented 9 years ago

Turns out we can build a recipe for a package we need and then build a conda package from it. Next, we upload the build to binstar, so we can create an environment directly from the build. I started with a recipe for protpy, which is required for PCM workflow and pushed it to a new repo called conda-recipes. I'll create an organization account on binstar to keep all the build used in our projects. More on this later.

chaninn commented 9 years ago

Can you name the organization cdmbi

Thanks

On Sunday, March 29, 2015, Likit Preeyanon notifications@github.com wrote:

Turns out we can build a recipe for a package we need and then build a conda package from it. Next, we upload the build to binstar, so we create an environment directly from it. I started with a recipe for protpy, which is required for PCM workflow and pushed it to a new repo called conda-recipes. I'll create an organization account on binstar to keep all the build used in our projects. More on this later.

— Reply to this email directly or view it on GitHub https://github.com/cdmbi/PCM/issues/2#issuecomment-87399727.

Associate Professor Chanin Nantasenamat, Ph.D. Head, Center of Data Mining and Biomedical Informatics Faculty of Medical Technology, Mahidol University 999 Phutthamonthon 4 Road, Salaya, Nakhon Pathom 73170 Tel. 02 441 4371 ext 2715 Fax 02 441 4380

likit commented 9 years ago

Definitely.

chaninn commented 9 years ago

Great.

On Sunday, March 29, 2015, Likit Preeyanon notifications@github.com wrote:

Definitely.

— Reply to this email directly or view it on GitHub https://github.com/cdmbi/PCM/issues/2#issuecomment-87404442.

Associate Professor Chanin Nantasenamat, Ph.D. Head, Center of Data Mining and Biomedical Informatics Faculty of Medical Technology, Mahidol University 999 Phutthamonthon 4 Road, Salaya, Nakhon Pathom 73170 Tel. 02 441 4371 ext 2715 Fax 02 441 4380