PineForge HPO 1.11.0
Native hyperparameter optimization for PineForge strategies
Loading...
Searching...
No Matches
pineforge_hpo.neighbourhood.Study Class Reference

Public Member Functions

None __init__ (self, list sources, Any passed_space=None)
 
str key_of (self, dict values)
 
dict row_values (self, Row row)
 
None bind (self, dict band, float|None fraction, float|None ratio)
 
float gap (self, float objective)
 
bool better (self, float objective, float other)
 
Any rank (self, str name, Any value)
 
list sort_values (self, str name, Any values)
 
dict cell (self, dict values)
 

Public Attributes

 space
 
 direction
 
 dimensions
 
 names
 
 varying
 
 constants
 
 rows
 
 row_count
 
 by_id
 
 summaries
 
 input_sha256
 
 duplicate_rows
 
 reference
 
 reference_id
 
 band
 
 delta
 
 fraction
 
 ratio
 

Detailed Description

The rows of the supplied sources indexed by parameter vector.

Built by ``prepare``. It holds the common recorded space, one ``Dimension``
per parameter in byte-wise name order, the groups of rows that share a
vector, the reference value and the input digest. ``bind`` adds the band so
that ``cell`` can classify a vector. Treat every attribute as read-only.

Definition at line 493 of file neighbourhood.py.

Constructor & Destructor Documentation

◆ __init__()

None pineforge_hpo.neighbourhood.Study.__init__ (   self,
list  sources,
Any   passed_space = None 
)

Definition at line 502 of file neighbourhood.py.

Member Function Documentation

◆ better()

bool pineforge_hpo.neighbourhood.Study.better (   self,
float  objective,
float  other 
)

◆ bind()

None pineforge_hpo.neighbourhood.Study.bind (   self,
dict  band,
float | None  fraction,
float | None  ratio 
)
Set the band and the move rules; computes the near-best distance.

Definition at line 599 of file neighbourhood.py.

References pineforge_hpo.neighbourhood.Study.reference.

◆ cell()

dict pineforge_hpo.neighbourhood.Study.cell (   self,
dict  values 
)
Class, objective, gap, near flag and row ids of one full vector.

Classes: ``scored_ok`` (status ok), ``scored_infeasible`` (constraint
violation), ``unscorable`` (any other status; the recorded objective of
the first row is shown, a pruned row's is a prefix value) and
``never_evaluated``. Scored rows of one vector agree exactly, so the
first scored row speaks for the vector; otherwise the first row does.

Definition at line 632 of file neighbourhood.py.

References pineforge_hpo.neighbourhood.Study.delta, pineforge_hpo.neighbourhood.Study.gap(), and pineforge_hpo.neighbourhood.Study.key_of().

◆ gap()

float pineforge_hpo.neighbourhood.Study.gap (   self,
float  objective 
)

◆ key_of()

str pineforge_hpo.neighbourhood.Study.key_of (   self,
dict  values 
)
Canonical key of a normalised parameter vector.

Definition at line 548 of file neighbourhood.py.

References pineforge_hpo._rows.Row.space, pineforge_hpo._rows.Source.space, and pineforge_hpo.neighbourhood.Study.space.

Referenced by pineforge_hpo.neighbourhood.Study.cell().

◆ rank()

Any pineforge_hpo.neighbourhood.Study.rank (   self,
str  name,
Any  value 
)

◆ row_values()

dict pineforge_hpo.neighbourhood.Study.row_values (   self,
Row  row 
)
The normalised parameter vector of a row, or an input_invalid refusal.

Definition at line 555 of file neighbourhood.py.

References pineforge::hpo::SearchSpace.dimensions(), and pineforge_hpo.neighbourhood.Study.dimensions.

◆ sort_values()

list pineforge_hpo.neighbourhood.Study.sort_values (   self,
str  name,
Any  values 
)

Member Data Documentation

◆ band

pineforge_hpo.neighbourhood.Study.band

Definition at line 609 of file neighbourhood.py.

◆ by_id

pineforge_hpo.neighbourhood.Study.by_id

Definition at line 521 of file neighbourhood.py.

◆ constants

pineforge_hpo.neighbourhood.Study.constants

Definition at line 518 of file neighbourhood.py.

◆ delta

pineforge_hpo.neighbourhood.Study.delta

Definition at line 610 of file neighbourhood.py.

Referenced by pineforge_hpo.neighbourhood.Study.cell().

◆ dimensions

◆ direction

pineforge_hpo.neighbourhood.Study.direction

◆ duplicate_rows

pineforge_hpo.neighbourhood.Study.duplicate_rows

Definition at line 530 of file neighbourhood.py.

◆ fraction

pineforge_hpo.neighbourhood.Study.fraction

Definition at line 611 of file neighbourhood.py.

◆ input_sha256

pineforge_hpo.neighbourhood.Study.input_sha256

Definition at line 529 of file neighbourhood.py.

◆ names

◆ ratio

pineforge_hpo.neighbourhood.Study.ratio

Definition at line 612 of file neighbourhood.py.

◆ reference

pineforge_hpo.neighbourhood.Study.reference

◆ reference_id

pineforge_hpo.neighbourhood.Study.reference_id

Definition at line 540 of file neighbourhood.py.

◆ row_count

pineforge_hpo.neighbourhood.Study.row_count

Definition at line 520 of file neighbourhood.py.

◆ rows

pineforge_hpo.neighbourhood.Study.rows

Definition at line 519 of file neighbourhood.py.

◆ space

pineforge_hpo.neighbourhood.Study.space

Definition at line 505 of file neighbourhood.py.

Referenced by pineforge_hpo.neighbourhood.Study.key_of().

◆ summaries

pineforge_hpo.neighbourhood.Study.summaries

Definition at line 522 of file neighbourhood.py.

◆ varying

pineforge_hpo.neighbourhood.Study.varying

Definition at line 517 of file neighbourhood.py.


The documentation for this class was generated from the following file: