You will have to install the Atomic Simulation Environment for this.

! pip install ase

We will use it because it has data about the chemical elements, and some molecules we will use.

First, we make a DataFrame containing the chemical elements, and their atomic masses.

import ase
import numpy as np
import pandas as pd

dtypes = np.dtype([('symbol', str), ('atomic mass', float)])
data = np.empty(0, dtype=dtypes)

elements = pd.DataFrame(data)
elements['symbol'] = ase.data.chemical_symbols
elements['atomic mass'] = ase.data.atomic_masses
elements.info()

<class 'pandas.core.frame.DataFrame'> RangeIndex: 119 entries, 0 to 118 Data columns (total 2 columns):

— -–— ---------–— –— 0 symbol 119 non-null object 1 atomic mass 119 non-null float64 dtypes: float64(1), object(1) memory usage: 2.0+ KB

You can use a query function to select rows from the database that meet some criteria.

?elements.query

For example, to get Carbon we can use a query like this.

elements.query('symbol == "C"')

  symbol  atomic mass
6      C       12.011

We can do some things with python variables like this.

sym = 'H'
elements.query('symbol == @sym')

  symbol  atomic mass
1      H        1.008

#+END_SRC

Although, f-strings also work for this.

sym = 'H'
elements.query(f'symbol == "{sym}"')

  symbol  atomic mass
1      H        1.008

The atomic mass is trickier to work with because it has a space in it. We have to use back-ticks to "quote" this.

elements.query('`atomic mass` < 4')

  symbol  atomic mass
0      X        1.000
1      H        1.008

Note it does not appear you can use the @ or f-string on column names. Also note the mysterious element X in the table. We can ignore that.

Next, let's build the molecule database. This will be a table where each row corresponds to an atom in a molecule. We will be able to get the atoms in a molecule by aggregating these on a Molecule-ID. We build this table up row by row, by first getting a molecule, and then iterating over each atom in the molecule. It is conventional to use an integer for an ID, but we will just use the molecular formula in this example. That has some limitations for larger databases (e.g. isomers have different properties, but the same molecular formula), but we will not have that problem here.

df = pd.DataFrame(columns=['Molecule-ID', 'Atom symbol', 'x', 'y', 'z'])

from ase.build import molecule
i = 0

for mlc in ['H2O', 'NH3', 'CH4']:
    for atom in molecule(mlc):
        df.loc[i] = [mlc, atom.symbol, atom.x, atom.y, atom.z]
        i += 1
df

   Molecule-ID Atom symbol         x         y         z
0          H2O           O  0.000000  0.000000  0.119262
1          H2O           H  0.000000  0.763239 -0.477047
2          H2O           H  0.000000 -0.763239 -0.477047
3          NH3           N  0.000000  0.000000  0.116489
4          NH3           H  0.000000  0.939731 -0.271808
5          NH3           H  0.813831 -0.469865 -0.271808
6          NH3           H -0.813831 -0.469865 -0.271808
7          CH4           C  0.000000  0.000000  0.000000
8          CH4           H  0.629118  0.629118  0.629118
9          CH4           H -0.629118 -0.629118  0.629118
10         CH4           H  0.629118 -0.629118 -0.629118
11         CH4           H -0.629118  0.629118 -0.629118

Getting the molecular weight requires us to combine information from two DataFrames. To do this, we need to merge them, aligning the rows on a common key. That key is the Atom symbol in the molecule DataFrame, and symbol in the elements DataFrame. Then we have to do the right grouping, and use the sum aggregation method on each group.

mf = pd.merge(df, elements, how='inner', left_on='Atom symbol', right_on='symbol')
mf

   Molecule-ID Atom symbol         x         y         z symbol  atomic mass
0          H2O           O  0.000000  0.000000  0.119262      O       15.999
1          H2O           H  0.000000  0.763239 -0.477047      H        1.008
2          H2O           H  0.000000 -0.763239 -0.477047      H        1.008
3          NH3           H  0.000000  0.939731 -0.271808      H        1.008
4          NH3           H  0.813831 -0.469865 -0.271808      H        1.008
5          NH3           H -0.813831 -0.469865 -0.271808      H        1.008
6          CH4           H  0.629118  0.629118  0.629118      H        1.008
7          CH4           H -0.629118 -0.629118  0.629118      H        1.008
8          CH4           H  0.629118 -0.629118 -0.629118      H        1.008
9          CH4           H -0.629118  0.629118 -0.629118      H        1.008
10         NH3           N  0.000000  0.000000  0.116489      N       14.007
11         CH4           C  0.000000  0.000000  0.000000      C       12.011

Now, we group by the Molecule-ID, select the atomic mass column, and aggregate with the sum.

MW = mf.groupby('Molecule-ID')['atomic mass'].sum()
MW

Molecule-ID
CH4    16.043
H2O    18.015
NH3    17.031
Name: atomic mass, dtype: float64

Here is one of many ways to print this in a different format:

for mlc, mw in MW.iteritems():
    print(f'The molecular weight of {mlc} is {mw} gm/mol')

The molecular weight of CH4 is 16.043 gm/mol The molecular weight of H2O is 18.015 gm/mol The molecular weight of NH3 is 17.031 gm/mol