[Tutor] Euclidean Distances between Atoms in a Molecule.

[Stephen P. Molnar]

On 04/03/2017 05:36 AM, Peter Otten wrote:
> Stephen P. Molnar wrote:
>
>> I am trying to port a program that I wrote in FORTRAN twenty years ago
>> into Python 3 and am having a hard time trying to calculate the
>> Euclidean distance between each atom in the molecule and every other
>> atom in the molecule.
>>
>> Here is a typical table of coordinates:
>>
>>
>>         MASS         X         Y         Z
>> 0   12.011 -3.265636  0.198894  0.090858
>> 1   12.011 -1.307161  1.522212  1.003463
>> 2   12.011  1.213336  0.948208 -0.033373
>> 3   14.007  3.238650  1.041523  1.301322
>> 4   12.011 -5.954489  0.650878  0.803379
>> 5   12.011  5.654476  0.480066  0.013757
>> 6   12.011  6.372043  2.731713 -1.662411
>> 7   12.011  7.655753  0.168393  2.096802
>> 8   12.011  5.563051 -1.990203 -1.511875
>> 9    1.008 -2.939469 -1.327967 -1.247635
>> 10   1.008 -1.460475  2.993912  2.415410
>> 11   1.008  1.218042  0.451815 -2.057439
>> 12   1.008 -6.255901  2.575035  1.496984
>> 13   1.008 -6.560562 -0.695722  2.248982
>> 14   1.008 -7.152500  0.390758 -0.864115
>> 15   1.008  4.959548  3.061356 -3.139100
>> 16   1.008  8.197613  2.429073 -2.588339
>> 17   1.008  6.503322  4.471092 -0.543939
>> 18   1.008  7.845274  1.892126  3.227577
>> 19   1.008  9.512371 -0.273198  1.291080
>> 20   1.008  7.147039 -1.365346  3.393778
>> 21   1.008  4.191488 -1.928466 -3.057804
>> 22   1.008  5.061650 -3.595015 -0.302810
>> 23   1.008  7.402586 -2.392148 -2.374554
>>
>> What I need for further calculation is a matrix of the Euclidean
>> distances between the atoms.
>>
>> So far in searching the Python literature I have only managed to confuse
>> myself and would greatly appreciate any pointers towards a solution.
>>
>> Thanks in advance.
>>
>
> Stitched together with heavy use of a search engine:
>
> $ cat data.txt
>         MASS         X         Y         Z
> 0   12.011 -3.265636  0.198894  0.090858
> 1   12.011 -1.307161  1.522212  1.003463
> 2   12.011  1.213336  0.948208 -0.033373
> 3   14.007  3.238650  1.041523  1.301322
> 4   12.011 -5.954489  0.650878  0.803379
> 5   12.011  5.654476  0.480066  0.013757
> 6   12.011  6.372043  2.731713 -1.662411
> 7   12.011  7.655753  0.168393  2.096802
> 8   12.011  5.563051 -1.990203 -1.511875
> 9    1.008 -2.939469 -1.327967 -1.247635
> 10   1.008 -1.460475  2.993912  2.415410
> 11   1.008  1.218042  0.451815 -2.057439
> 12   1.008 -6.255901  2.575035  1.496984
> 13   1.008 -6.560562 -0.695722  2.248982
> 14   1.008 -7.152500  0.390758 -0.864115
> 15   1.008  4.959548  3.061356 -3.139100
> 16   1.008  8.197613  2.429073 -2.588339
> 17   1.008  6.503322  4.471092 -0.543939
> 18   1.008  7.845274  1.892126  3.227577
> 19   1.008  9.512371 -0.273198  1.291080
> 20   1.008  7.147039 -1.365346  3.393778
> 21   1.008  4.191488 -1.928466 -3.057804
> 22   1.008  5.061650 -3.595015 -0.302810
> 23   1.008  7.402586 -2.392148 -2.374554
> $ python3
> Python 3.4.3 (default, Nov 17 2016, 01:08:31)
> [GCC 4.8.4] on linux
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import numpy, pandas, scipy.spatial.distance as dist
>>>> df = pandas.read_table("data.txt", sep=" ", skipinitialspace=True)
>>>> a = numpy.array(df[["X", "Y", "Z"]])
>>>> dist.squareform(dist.pdist(a, "euclidean"))
> <snip big matrix>
>
> Here's an example with just the first 4 atoms:
>
>>>> dist.squareform(dist.pdist(a[:4], "euclidean"))
> array([[ 0.        ,  2.53370139,  4.54291701,  6.6694065 ],
>         [ 2.53370139,  0.        ,  2.78521357,  4.58084922],
>         [ 4.54291701,  2.78521357,  0.        ,  2.42734737],
>         [ 6.6694065 ,  4.58084922,  2.42734737,  0.        ]])
>
> See
> https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.distance.pdist.html
> There may be a way to do this with pandas.pivot_table(), but I didn't manage
> to find that.
>
> As Alan says, this is not the appropriate forum for the topic you are
> belabouring.
>
> Work your way through a Python tutorial to pick up the basics (we can help
> you with this), then go straight to where the (numpy/scipy) experts are.
>
> _______________________________________________
> Tutor maillist  -  Tutor at python.org
> To unsubscribe or change subscription options:
> https://mail.python.org/mailman/listinfo/tutor
>

Thanks to everyone who answered.

The problem has been solved.

-- 
Stephen P. Molnar, Ph.D.		Life is a fuzzy set
www.molecular-modeling.net		Stochastic and multivariate
(614)312-7528 (c)
Skype: smolnar1