Sorry, maybe this is not very related to this phase of this kaggle but it is a great help of you if you assist me to solve this probelm\nI have a problem to activate both rdkit and tensorflow in anaconda. When I activate one of them the another is deactivate and so on! Do you have any suggestion?
", "messages": [ { "id": "607184", "postDate": "08/24/2019 18:25:58", "content": "Sorry, maybe this is not very related to this phase of this kaggle but it is a great help of you if you assist me to solve this probelm\nI have a problem to activate both rdkit and tensorflow in anaconda. When I activate one of them the another is deactivate and so on! Do you have any suggestion?
", "rawMarkdown": "Sorry, maybe this is not very related to this phase of this kaggle but it is a great help of you if you assist me to solve this probelm\nI have a problem to activate both rdkit and tensorflow in anaconda. When I activate one of them the another is deactivate and so on! Do you have any suggestion?", "votes": null }, { "id": "607201", "postDate": "08/24/2019 19:19:29", "content": "Just did this last night - I used PIP - wanted GPU - so tensorflow-gpu and keras-gpu. Install order was RDkit - tensorflow - keras - this is on Windows 10 system. I first did uninstall of tensorflow and keras because like you rdkit wanted to remove tensorflow.
", "rawMarkdown": "Just did this last night - I used PIP - wanted GPU - so tensorflow-gpu and keras-gpu. Install order was RDkit - tensorflow - keras - this is on Windows 10 system. I first did uninstall of tensorflow and keras because like you rdkit wanted to remove tensorflow.", "votes": null }, { "id": "607228", "postDate": "08/24/2019 21:05:49", "content": "I don't know why you need both in a single kernel/environment, but if you use Rdkit for preparing data and tensorflow for machine learning you can do it in separate kernels/environments. In a kernel, it is possible to use output data from another kernel.
", "rawMarkdown": "I don't know why you need both in a single kernel/environment, but if you use Rdkit for preparing data and tensorflow for machine learning you can do it in separate kernels/environments. In a kernel, it is possible to use output data from another kernel.", "votes": null }, { "id": "607285", "postDate": "08/25/2019 00:44:29", "content": "rdkit and tf needs Python 3.6. It is highly likely you are on Python 3.7 anaconda dist if you keep it up-to-date automatically. The solution is to create a virtual env. When you need to run them, activate the virtual env.
\n\n\nWhen I activate one of them the another is deactivate
\n
It looks like you installed rdkit and tf in two different conda environments. You should install them in the same environment. And, as Shuhao Cao writes, create a conda env with python 3.6 first.
", "rawMarkdown": "> When I activate one of them the another is deactivate\n\nIt looks like you installed rdkit and tf in two different conda environments. You should install them in the same environment. And, as Shuhao Cao writes, create a conda env with python 3.6 first.", "votes": null } ], "comments": [ { "id": 607201, "author_name": "pcjimmmy", "author_url": "", "post_date": "08/24/2019 19:19:29", "content": "Just did this last night - I used PIP - wanted GPU - so tensorflow-gpu and keras-gpu. Install order was RDkit - tensorflow - keras - this is on Windows 10 system. I first did uninstall of tensorflow and keras because like you rdkit wanted to remove tensorflow.
", "votes": null, "replies": [] }, { "id": 607228, "author_name": "somnambwl", "author_url": "", "post_date": "08/24/2019 21:05:49", "content": "I don't know why you need both in a single kernel/environment, but if you use Rdkit for preparing data and tensorflow for machine learning you can do it in separate kernels/environments. In a kernel, it is possible to use output data from another kernel.
", "votes": null, "replies": [] }, { "id": 607285, "author_name": "scaomath", "author_url": "", "post_date": "08/25/2019 00:44:29", "content": "rdkit and tf needs Python 3.6. It is highly likely you are on Python 3.7 anaconda dist if you keep it up-to-date automatically. The solution is to create a virtual env. When you need to run them, activate the virtual env.
\n\n\nWhen I activate one of them the another is deactivate
\n
It looks like you installed rdkit and tf in two different conda environments. You should install them in the same environment. And, as Shuhao Cao writes, create a conda env with python 3.6 first.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "607184": "Sorry, maybe this is not very related to this phase of this kaggle but it is a great help of you if you assist me to solve this probelm\nI have a problem to activate both rdkit and tensorflow in anaconda. When I activate one of them the another is deactivate and so on! Do you have any suggestion?", "607201": "Just did this last night - I used PIP - wanted GPU - so tensorflow-gpu and keras-gpu. Install order was RDkit - tensorflow - keras - this is on Windows 10 system. I first did uninstall of tensorflow and keras because like you rdkit wanted to remove tensorflow.", "607228": "I don't know why you need both in a single kernel/environment, but if you use Rdkit for preparing data and tensorflow for machine learning you can do it in separate kernels/environments. In a kernel, it is possible to use output data from another kernel.", "607285": "`rdkit` and `tf` needs Python 3.6. It is highly likely you are on Python 3.7 anaconda dist if you keep it up-to-date automatically. The solution is to create a virtual env. When you need to run them, activate the virtual env.", "607329": "> When I activate one of them the another is deactivate\n\nIt looks like you installed rdkit and tf in two different conda environments. You should install them in the same environment. And, as Shuhao Cao writes, create a conda env with python 3.6 first." }, "source": "meta" }