rxnutils.pipeline package¶
Subpackages¶
- rxnutils.pipeline.actions package
- Submodules
- rxnutils.pipeline.actions.dataframe_mod module
- rxnutils.pipeline.actions.reaction_mod module
DesaltMolecules
NameRxn
NeutralizeMolecules
ReactantsToReagents
ReagentsToReactants
RemoveAtomMapping
RemoveStereoInfo
InvertStereo
IsotopeInfo
RemoveExtraAtomMapping
RemoveUnchangedProducts
RemoveUnsanitizable
RDKitRxnRoles
RxnMapper
SplitReaction
AtomMapTagDisconnectionSite
ConvertAtomMapDisconnectionTag
TrimRxnSmiles
- rxnutils.pipeline.actions.reaction_props module
CountComponents
CountElements
HasStereoInfo
HasUnmappedRadicalAtom
HasUnsanitizableReactants
CgrCreated
CgrNumberOfDynamicBonds
ProductAtomMappingStats
ProductSize
PseudoReactionHash
PseudoSmilesHash
ReactantProductAtomBalance
ReactantSize
MaxRingNumber
RingNumberChange
RingBondMade
RingMadeSize
SmilesLength
SmilesSanitizable
StereoInvention
StereoCentreChanges
StereoHasChiralReagent
StereoCenterIsCreated
StereoCenterIsRemoved
StereoCenterInReactantPotential
StereoCenterOutsideReaction
StereoMesoProduct
- rxnutils.pipeline.actions.templates module
- Module contents
Submodules¶
rxnutils.pipeline.base module¶
Module containing routines for the validation framework
- rxnutils.pipeline.base.action(obj)¶
Decorator that register a callable as a validation action.
An action will be called with a pandas.DataFrame object and return a new pandas.DataFrame object.
An action needs to have an attribute pretty_name.
- Parameters:
obj (Callable[[DataFrame], DataFrame]) – the callable to register as an action
- Returns:
the same as obj.
- Return type:
Callable[[DataFrame], DataFrame]
- rxnutils.pipeline.base.list_actions(short=False)¶
List all available actions in a nice table
- Parameters:
short (bool)
- Return type:
None
- rxnutils.pipeline.base.create_action(pretty_name, *args, **kwargs)¶
Create an action that can be called
- Parameters:
pretty_name (str) – the name of the action
args (Any)
kwargs (Any)
- Returns:
the instantiated actions
- Return type:
Callable[[DataFrame], DataFrame]
- class rxnutils.pipeline.base.ReactionActionMixIn¶
Bases:
object
Mixin class with standard routines for splitting and joining reaction SMILES
- join_lists(reactants_list, reagents_list, products_list)¶
Join list of components into a reaction SMILES
- Parameters:
reactants_list (List[str]) – the list of reactant SMILES
reagents_list (List[str]) – the list of reagent SMILES
products_list (List[str]) – the list of product SMILES
- Returns:
the concatenated reaction SMILES
- Return type:
str
- join_smiles(reactants, reagents, products)¶
Join component SMILES into a reaction SMILES
- Parameters:
reactants (str) – the reactant SMILES
reagents (str) – the reagent SMILES
products (str) – the product SMILES
- Returns:
the concatenated reaction SMILES
- Return type:
str
- split_lists(row)¶
Split a reaction SMILES into list of component SMILES
- Parameters:
row (Series) – the row with the SMILES
- Returns:
the list of SMILES of the components
- Return type:
Tuple[List[str], List[str], List[str]]
- split_smiles(row)¶
Split a reaction SMILES into components SMILES
- Parameters:
row (Series) – the row with the SMILES
- Returns:
the SMILES of the components
- Return type:
Tuple[str, str, str]
rxnutils.pipeline.runner module¶
Module containg routines and interface to run pipelines
- rxnutils.pipeline.runner.run_pipeline(data, pipeline, filename, save_intermediates=True)¶
Run a given pipeline on a dataset
The actions are applied sequentials as they are defined in the pipeline
The intermediate results of the pipeline will be written to separate tab-separated CSV files.
- Parameters:
data (DataFrame) – the dataset
pipeline (Dict[str, Any]) – the action specifications
filename (str) – path to the final output file
save_intermediates (bool) – if True will save intermediate results
- Returns:
the dataset after completing the pipeline
- Return type:
DataFrame
- rxnutils.pipeline.runner.main(args=None)¶
Function for command line argument
- Parameters:
args (Sequence[str] | None)
- Return type:
None