You can run this notebook online in a Binder session or view it on Github.

Exploring Datasets and ReactionDatasets

When examining a Dataset or ReactionDataset, it can be helpful to see which fields and associated values are available. That is, this section will discuss how to list what basis sets, methods, programs, etc. are represented in a dataset’s results.

import qcportal as ptl
client = ptl.FractalClient()

ds = client.get_collection("ReactionDataset", "S22")

To see available search parameters:

ds.list_values().reset_index().columns

Index(['native', 'driver', 'program', 'method', 'basis', 'keywords',
       'stoichiometry', 'name'],
      dtype='object')

To see available search values (e.g. which basis sets and methods are available in the dataset):

ds.list_values().reset_index()['method'].unique()

array(['Unknown', 'b2plyp', 'b2plyp-d3', 'b2plyp-d3(bj)', 'b2plyp-d3m',
       'b2plyp-d3m(bj)', 'b3lyp', 'b3lyp-d3', 'b3lyp-d3(bj)', 'b3lyp-d3m',
       'b3lyp-d3m(bj)', 'hf', 'mp2', 'pbe', 'sapt0', 'wb97m-v', 'wb97x-d'],
      dtype=object)

ds.list_values().reset_index()['basis'].unique()

array(['Unknown', 'aug-cc-pvdz', 'aug-cc-pvtz', 'def2-svp', 'def2-tzvp',
       'sto-3g', 'jun-cc-pvdz'], dtype=object)

List combinations of method and basis set:

ds.list_values().reset_index().groupby(['method','basis']).size().reset_index()

	method	basis	0
0	Unknown	Unknown	3
1	b2plyp	aug-cc-pvdz	2
2	b2plyp	aug-cc-pvtz	2
3	b2plyp	def2-svp	2
4	b2plyp	def2-tzvp	2
5	b2plyp-d3	aug-cc-pvdz	2
6	b2plyp-d3	aug-cc-pvtz	2
7	b2plyp-d3	def2-svp	2
8	b2plyp-d3	def2-tzvp	2
9	b2plyp-d3(bj)	aug-cc-pvdz	2
10	b2plyp-d3(bj)	aug-cc-pvtz	2
11	b2plyp-d3(bj)	def2-svp	2
12	b2plyp-d3(bj)	def2-tzvp	2
13	b2plyp-d3m	aug-cc-pvdz	2
14	b2plyp-d3m	aug-cc-pvtz	2
15	b2plyp-d3m	def2-svp	2
16	b2plyp-d3m	def2-tzvp	2
17	b2plyp-d3m(bj)	aug-cc-pvdz	2
18	b2plyp-d3m(bj)	aug-cc-pvtz	2
19	b2plyp-d3m(bj)	def2-svp	2
20	b2plyp-d3m(bj)	def2-tzvp	2
21	b3lyp	aug-cc-pvdz	2
22	b3lyp	aug-cc-pvtz	2
23	b3lyp	def2-svp	2
24	b3lyp	def2-tzvp	2
25	b3lyp-d3	aug-cc-pvdz	2
26	b3lyp-d3	aug-cc-pvtz	2
27	b3lyp-d3	def2-svp	2
28	b3lyp-d3	def2-tzvp	2
29	b3lyp-d3(bj)	aug-cc-pvdz	2
30	b3lyp-d3(bj)	aug-cc-pvtz	2
31	b3lyp-d3(bj)	def2-svp	2
32	b3lyp-d3(bj)	def2-tzvp	2
33	b3lyp-d3m	aug-cc-pvdz	2
34	b3lyp-d3m	aug-cc-pvtz	2
35	b3lyp-d3m	def2-svp	2
36	b3lyp-d3m	def2-tzvp	2
37	b3lyp-d3m(bj)	aug-cc-pvdz	2
38	b3lyp-d3m(bj)	aug-cc-pvtz	2
39	b3lyp-d3m(bj)	def2-svp	2
40	b3lyp-d3m(bj)	def2-tzvp	2
41	hf	sto-3g	2
42	mp2	aug-cc-pvdz	2
43	mp2	aug-cc-pvtz	2
44	mp2	def2-svp	2
45	mp2	def2-tzvp	2
46	pbe	aug-cc-pvdz	2
47	pbe	aug-cc-pvtz	2
48	pbe	def2-svp	2
49	pbe	def2-tzvp	2
50	sapt0	aug-cc-pvdz	1
51	sapt0	aug-cc-pvtz	1
52	sapt0	jun-cc-pvdz	1
53	wb97m-v	def2-svp	2
54	wb97m-v	def2-tzvp	2
55	wb97x-d	def2-svp	2
56	wb97x-d	def2-tzvp	2

Most datasets support default programs and keywords. For those cases, the above can also be achieved with:

ds.list_values(native=True).reset_index()['name']

0      cp-B2PLYP/aug-cc-pvdz
1         B2PLYP/aug-cc-pvdz
2      cp-B2PLYP/aug-cc-pvtz
3         B2PLYP/aug-cc-pvtz
4            B2PLYP/def2-svp
               ...
104        WB97M-V/def2-tzvp
105      cp-WB97X-D/def2-svp
106         WB97X-D/def2-svp
107        WB97X-D/def2-tzvp
108     cp-WB97X-D/def2-tzvp
Name: name, Length: 109, dtype: object