02-批量汇总、过滤与组合¶

In [1]:

from __future__ import annotations

from pathlib import Path
from molop.io import AutoParser


def find_repo_root(start: Path) -> Path:
    current = start.resolve()
    for candidate in (current, *current.parents):
        if (candidate / "pyproject.toml").exists():
            return candidate
    raise RuntimeError("Could not find repository root (pyproject.toml).")


repo_root = find_repo_root(Path.cwd())

from __future__ import annotations from pathlib import Path from molop.io import AutoParser def find_repo_root(start: Path) -> Path: current = start.resolve() for candidate in (current, *current.parents): if (candidate / "pyproject.toml").exists(): return candidate raise RuntimeError("Could not find repository root (pyproject.toml).") repo_root = find_repo_root(Path.cwd())

批量汇总¶

In [2]:

mix_batch = AutoParser((repo_root / "tests/test_files/mix_format/*").as_posix())
mix_batch

mix_batch = AutoParser((repo_root / "tests/test_files/mix_format/*").as_posix()) mix_batch

MolOP parsing with single process:   0%|          | 0/10 [00:00<?, ?it/s]

Out[2]:

FileBatchModelDisk(10)

In [33]:

mix_batch.to_summary_df()

mix_batch.to_summary_df()

MolOP processing frame summary with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[33]:

	DiskStorage		General					Calc Parameter						Environment				Status
	FilePath	FileFormat	Charge	Multiplicity	CanonicalSMILES	NumAtoms	FrameID	Software	Version	Method	...	Functional	Keywords	SolventModel	Solvent	Temperature (kelvin)	Pressure (standard_atmosphere)	IsError	IsNormal	IsTS	IsOptimized
0	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.log	-1	1	C/C=C(\[O-])[N-](c1ccccc1)->[Cu+3]1(<-[O-]c2cc...	77	131	Gaussian	ES64L-G16RevC.01	DFT	...	RB3LYP	#N Geom=AllCheck Guess=TCheck SCRF=Check GenCh...	None	None	298.15 kelvin	1.0 standard_atmosphere	False	True	True	False
1	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.gjf	0	1	CCOC(=O)c1cc(OC)no1	21	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.gjf	0	1	C=[N+](C)[N-]C.COC(=O)[C@@]1(OC)C#CC(Br)(Br)CCCC1	43	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.log	0	1	C=[N+](C)[N-]C.COC(=O)[C@@]1(OC)C#CC(Br)(Br)CCCC1	43	73	Gaussian	ES64L-G16RevB.01	DFT	...	RB3LYP	#N Geom=AllCheck Guess=TCheck SCRF=Check Test ...	PCM	Water	298.15 kelvin	1.0 standard_atmosphere	False	True	True	False
4	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.fchk	0	2	[CH2]C	7	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
5	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.gjf	0	2	[CH2]C	7	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
6	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.log	1	1	C[N+]1=C=CC([N+](=O)[O-])=C1	14	21	Gaussian	ES64L-G16RevC.01	DFT	...	RB3LYP-GD3BJ	#p b3lyp/chkbas pop=nbo geom=allcheck guess=tc...	None	None	298.15 kelvin	1.0 standard_atmosphere	False	True	False	False
7	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.gjf	0	1	C=CCC.CC(=O)O	20	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
8	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.out	0	1	C=CCC.CC(=O)O	20	43	Gaussian	Gaussian 09: AS64L-G09RevD.01 24-Apr-2013\n ...	DFT	...		#P irc=(calcfc, MaxPoints=20) ub3lyp/6-31g(d)	None	None	NaN	NaN	True	False	False	False
9	/Users/tmj/Documents/proj/MolOP/tests/test_fil...	.gjf	0	1	C	5	0	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

10 rows × 21 columns

In [3]:

mix_batch.file_names

mix_batch.file_names

Out[3]:

['RE_BOX-Anion-Real_Cu-III-Phenol_Major-Amide-Anion_From-IP_C-O-190_TS_Opt.log',
 'S_Ph_Ni_TS.gjf',
 'TS_Zy0fwX_ll_ad_14-19_15-16_optts_g16.gjf',
 'TS_Zy0fwX_ll_ad_14-19_15-16_optts_g16.log',
 'dsgdb9nsd_000007-6.fchk',
 'dsgdb9nsd_000007-6.gjf',
 'dsgdb9nsd_131941-4+.log',
 'irc.gjf',
 'irc.out',
 'opt_point_charge_xtb.gjf']

可以像列表一样对批次进行索引，获取特定的文件。

In [4]:

mix_batch[0].file_path

mix_batch[0].file_path

Out[4]:

'/Users/tmj/Documents/proj/MolOP/tests/test_files/mix_format/RE_BOX-Anion-Real_Cu-III-Phenol_Major-Amide-Anion_From-IP_C-O-190_TS_Opt.log'

索引也可以是文件的路径，对于顺序不方便获得的大规模场景会更有用，例如：

In [5]:

mix_batch[mix_batch[0].file_path].file_path

mix_batch[mix_batch[0].file_path].file_path

Out[5]:

'/Users/tmj/Documents/proj/MolOP/tests/test_files/mix_format/RE_BOX-Anion-Real_Cu-III-Phenol_Major-Amide-Anion_From-IP_C-O-190_TS_Opt.log'

进行快速观察

In [6]:

mix_batch.draw_grid_image(molsPerRow=2, subImgSize=(500, 500))

mix_batch.draw_grid_image(molsPerRow=2, subImgSize=(500, 500))

MolOP processing smi format with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[6]:

如果计算结果中存在可嵌入到原子或键的信息，可以使用qm_embedded_rdmol方法获取嵌入后的rdkit mol对象，从这种对象保存的SDF/MOL文件中会包含这些信息。

In [7]:

mix_batch[0][-1].qm_embedded_rdmol()

mix_batch[0][-1].qm_embedded_rdmol()

Out[7]:

atom.dprop.MULLIKEN_CHARGES_BY_GAUSSIAN	0.118843 -0.52778700000000001 -0.51229000000000002 -0.53101200000000004 -0.51256599999999997 0.61099899999999996 0.0034429999999999999 0.139241 0.61233300000000002 0.115479 -0.39646199999999998 0.114855 0.0024759999999999999 0.125336 -0.057921 0.13408999999999999 0.12804299999999999 -0.50508900000000001 0.26441100000000001 0.111973 0.15039 0.122109 -0.45417600000000002 0.11734899999999999 0.13606699999999999 0.12819800000000001 -0.47093800000000002 0.11650099999999999 0.18501600000000001 0.137873 -0.065138000000000001 0.12690599999999999 0.15543299999999999 -0.476964 0.12060999999999999 0.120382 0.20754500000000001 -0.57657999999999998 -0.152752 -0.129075 -0.20485700000000001 0.41112599999999999 -0.20932500000000001 -0.14534900000000001 0.080785999999999997 0.084764000000000006 0.108117 0.16168299999999999 0.099557999999999994 -0.48423300000000002 0.120835 0.23432 0.13422200000000001 -0.45259300000000002 0.128138 0.13276499999999999 0.12720999999999999 0.55761000000000005 -0.52698400000000001 -0.57556099999999999 -0.084725999999999996 0.19952400000000001 -0.481348 0.178619 0.14002800000000001 0.18781600000000001 0.27968700000000002 -0.168215 -0.17993300000000001 -0.14568900000000001 0.17444000000000001 -0.141733 0.166348 -0.13463900000000001 0.098386000000000001 0.100982 0.091067999999999996
atom.dprop.APT_CHARGES_BY_GAUSSIAN	0.443965 -0.792883 -0.83162499999999995 -0.79886199999999996 -0.77464900000000003 1.2313970000000001 0.30693799999999999 -0.051416000000000003 1.167859 0.055232999999999997 -0.83615399999999995 0.040263 0.3377 -0.066701999999999997 0.38441500000000001 -0.050312000000000003 -0.081948999999999994 0.017576000000000001 0.048186 -0.056041000000000001 -0.0068040000000000002 0.062031000000000003 0.058993999999999998 -0.038065000000000002 -0.020482 -0.043746 0.028738 -0.045996000000000002 0.0098879999999999992 -0.010717000000000001 0.34742400000000001 -0.046920000000000003 -0.042823 0.038242999999999999 -0.031461000000000003 -0.038001 0.013105 -0.84482800000000002 -0.199792 0.099024000000000001 -0.18670300000000001 0.568415 -0.18141299999999999 0.061751 -0.039863999999999997 -0.049185 0.015035 0.047211999999999997 -0.015221 0.0093130000000000001 -0.053851999999999997 0.032781999999999999 -0.015025 0.050227000000000001 -0.050072999999999999 -0.025194000000000001 -0.027827999999999999 0.40771000000000002 -0.65421099999999999 -0.62746400000000002 0.75768800000000003 -0.028715000000000001 -0.16567699999999999 0.067944000000000004 -0.0081110000000000002 0.038531999999999997 0.388457 -0.182672 -0.149811 0.062414999999999998 0.099127000000000007 0.039350999999999997 0.047967000000000003 -0.142039 -0.032569000000000001 -0.012616 -0.026436000000000001

过滤文件¶

在文件数量较多时开启并行处理可以显著提高处理速度。

In [8]:

mix_batch.groupby(lambda x: x.pure_filename)

mix_batch.groupby(lambda x: x.pure_filename)

Grouping files with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[8]:

{'RE_BOX-Anion-Real_Cu-III-Phenol_Major-Amide-Anion_From-IP_C-O-190_TS_Opt': FileBatchModelDisk(1),
 'S_Ph_Ni_TS': FileBatchModelDisk(1),
 'TS_Zy0fwX_ll_ad_14-19_15-16_optts_g16': FileBatchModelDisk(2),
 'dsgdb9nsd_000007-6': FileBatchModelDisk(2),
 'dsgdb9nsd_131941-4+': FileBatchModelDisk(1),
 'irc': FileBatchModelDisk(2),
 'opt_point_charge_xtb': FileBatchModelDisk(1)}

In [24]:

mix_batch.groupby(lambda x: x.file_format)

mix_batch.groupby(lambda x: x.file_format)

Grouping files with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[24]:

{'.log': FileBatchModelDisk(3),
 '.gjf': FileBatchModelDisk(5),
 '.fchk': FileBatchModelDisk(1),
 '.out': FileBatchModelDisk(1)}

In [22]:

ts_batch = mix_batch.filter_state("ts")
ts_batch

ts_batch = mix_batch.filter_state("ts") ts_batch

Filtering ts files  with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[22]:

FileBatchModelDisk(2)

In [25]:

opt_batch = mix_batch.filter_state("opt")
opt_batch

opt_batch = mix_batch.filter_state("opt") opt_batch

Filtering opt files  with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[25]:

FileBatchModelDisk(3)

In [26]:

cation_batch = mix_batch.filter_value("charge", 1)
cation_batch

cation_batch = mix_batch.filter_value("charge", 1) cation_batch

Filtering files with charge == 1 with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[26]:

FileBatchModelDisk(1)

In [27]:

anion_batch = mix_batch.filter_value("charge", 0, "<")
anion_batch

anion_batch = mix_batch.filter_value("charge", 0, "<") anion_batch

Filtering files with charge < 0 with 1 jobs:   0%|          | 0/10 [00:00<?, ?it/s]

Out[27]:

FileBatchModelDisk(1)

内置了并行执行器，用户可以通过 parallel_execute 方法并行执行自定义函数，最终返回一个包含所有函数执行结果的列表。

In [28]:

mix_batch.parallel_execute(lambda x: print(x.charge))

mix_batch.parallel_execute(lambda x: print(x.charge))

  0%|          | 0/10 [00:00<?, ?it/s]

-1
0
0
0
0
0
1
0
0
0

Out[28]:

[None, None, None, None, None, None, None, None, None, None]

组合¶

支持简单的布尔运算，例如：

In [29]:

ion_batch = cation_batch + anion_batch
ion_batch

ion_batch = cation_batch + anion_batch ion_batch

Out[29]:

FileBatchModelDisk(2)

In [30]:

opt_batch_without_ts = opt_batch - ts_batch
opt_batch_without_ts

opt_batch_without_ts = opt_batch - ts_batch opt_batch_without_ts

Out[30]:

FileBatchModelDisk(1)