I just upgraded rdkit from 2017.03.1 to 2017.03.2 using Conda. What I
have noticed is that pandas are now installed during the installation
Does rdkit depend on pandas now? Is it safe to remove it? If it works
without pandas, maybe it makes sense to remove the dependency.
No. The main reason that the conda recipe includes pandas is for testing
the pandas extension. We could probably remove it from the run-time
dependency however and let the user install it in addition.
In any case, feel free to remove pandas from the conda installation.
Hello all ,
How about doing some dimension reduction using pca or Tsne and then run
clustering using some selected top components like top 20 and I think then
the clustering would be fast .
On Mon, Jun 5, 2017 at 6:11 AM David Cosgrove
On Tue, Jun 6, 2017 at 1:43 PM, Popov, Maxim (Ext) wrote:
> I have discovered a very usefule tool in Knime, Molecule Fragmenter by
> RDKit, but can’t find a corresponding class or function outside of Knime.
> Can I use the Fragmenter without Knime?
The KNIME node "RDKit Molecule Fragmenter" is indeed quite useful. Since what
it does for you in KNIME is based on RDKit functionality, you can do the same
also outside of KNIME. Usually, there is not a single method to call, rather a
code block that performs multiple operations,
I have discovered a very usefule tool in Knime, Molecule Fragmenter by RDKit,
but can't find a corresponding class or function outside of Knime. Can I use
the Fragmenter without Knime?
I was also searching for this functionality earlier
For what it’s worth here’s some *very* simple code I hacked together to do
fragmentation. The focus is aromatic heterocycles, but it could be more
general by, for example '[c,n]-[*]’ -> ‘[R]-[*]’ and ring =