banana.data package

Module to collect utilites to help with the data generation:

banana.data.cartesian_product(inp)[source]

Compute the cartesion product of a dictionary.

Thank you: https://stackoverflow.com/questions/5228158/

Parameters:: inp (dict) – dictionary listing all posibilities
Returns:: power set of all possibilities, preserving the keys
Return type:: list(dict)

Submodules

banana.data.db module

Actually using String it’s pointless, since SQLite has only one text type and it is TEXT, and the maximum length is coherently ignored by SQLite. Nevertheless the information is stored in the DB, even if the DBMS won’t use it, so we can keep declaring a size for fields where it matters.

Same considerations apply to datetime, with the further gain that SQLite it’s storing as he wish but able to use some dedicated functions.

class banana.data.db.Cache(**kwargs)[source]

Bases: CalcResult, Base

atime

ctime

external

hash

mtime

o_hash

pdf

result

t_hash

uid

class banana.data.db.CalcResult[source]

Bases: object

external = Column(None, Text(), table=None)

o_hash = Column(None, String(length=64), table=None)

pdf = Column(None, Text(), table=None)

t_hash = Column(None, String(length=64), table=None)

class banana.data.db.Log(**kwargs)[source]

Bases: CalcResult, Base

atime

ctime

external

hash

log

mtime

o_hash

pdf

t_hash

uid

class banana.data.db.MyBase[source]

Bases: object

atime = Column(None, DateTime(), table=None, onupdate=ColumnDefault(<function MyBase.<lambda>>))

ctime = Column(None, DateTime(), table=None, default=ColumnDefault(<function MyBase.<lambda>>))

hash = Column(None, String(length=64), table=None)

mtime = Column(None, DateTime(), table=None, onupdate=ColumnDefault(<function MyBase.<lambda>>))

uid = Column(None, Integer(), table=None, primary_key=True, nullable=False)

class banana.data.db.Theory(**kwargs)[source]

Bases: Base

CKM

Comments

DAMP

EScaleVar

FNS

GF

HQ

IB

IC

ID

MP

MW

MZ

MaxNfAs

MaxNfPdf

ModEv

NfFF

PTO

Q0

QED

Qedref

Qmb

Qmc

Qmt

Qref

SIN2TW

SxOrd

SxRes

TMC

XIF

XIR

alphaqed

alphas

atime

ctime

fact_to_ren_scale_ratio

global_nx

hash

kDISbThr

kDIScThr

kDIStThr

kbThr

kcThr

ktThr

mb

mc

mt

mtime

nfref

uid

banana.data.db.create_db(base_cls, engine)[source]

banana.data.db.engine(path='', dialect='sqlite', driver=None, username=None, password=None, host=None, port=None)[source]

banana.data.dfdict module

class banana.data.dfdict.DFdict(*args, **kwargs)[source]

Bases: dict

Collects dataframes in a dictionary printing them along side.

fancy(file=None)[source]

Print on stdout… with style!

Parameters:: file (IO[str]) – File to write to, None for stdout (passed down to rich)

classmethod from_document(d)[source]

Load dataframes from a previous dictionary dump

Parameters:: d (dict) – raw dictionary, containing a dump of a DFdict object
Returns:: the loaded DFdict
Return type:: DFdict

merge(other)[source]

print(*msgs, sep=' ', end='\n', position=None)[source]

Add new messages to the representation

Parameters:

msgs (str) – messages
sep (str) – separator
end (str) – end-of-line marker

q2_slice(q2)[source]

property q2s

to_document()[source]

Convert dataframes back to a true dictionary

Returns:: d – raw dictionary
Return type:: dict

banana.data.sql module

exception banana.data.sql.HashError(descr, objs=None)[source]

Bases: RetrieveError

objects = 'hashes'

exception banana.data.sql.RetrieveError(descr, objs=None)[source]

Bases: KeyError

objects = None

exception banana.data.sql.UIDError(descr, objs=None)[source]

Bases: RetrieveError

objects = 'pids'

banana.data.sql.add_hash(record)[source]

Add a hash value as last element to the record.

Parameters:: record (tuple) – data
Returns:: hashed_record – data + hash
Return type:: tuple

banana.data.sql.deserialize(data)[source]

Undo the binary representation.

Parameters:: data (list) – raw data
Returns:: obj – composed object
Return type:: dict

banana.data.sql.insertmany(session, table, df)[source]

Insert all records into the DB.

Parameters:

session (sqlalchemy.orm.session.Session) – database
table (sqlalchemy.schema.Table) – target table
df (pandas.DataFrame) – dataframe all records

banana.data.sql.insertnew(session, table, df)[source]

Insert all records that do not exist yet (determined by hash).

Parameters:

session (sqlalchemy.orm.session.Session) – database
table (sqlalchemy.schema.Table) – target table
df (pandas.DataFrame) – dataframe all records

banana.data.sql.prepare_records(base, updates)[source]

Generate all records from the base.

Parameters:

base (dict) – base record
updates (list(dict)) – update directives

Returns:

documents (list(dict)) – list of all dictionaries
df (pandas.DataFrame) – all records ready for insertion

banana.data.sql.remove(session, table_object, uids)[source]

Remove given rows from chosen table.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object
uids (list(int)) – unique identifiers of rows to delete

banana.data.sql.select_all(session, table_object)[source]

Collect all records.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object

Returns:

list of records

Return type:

banana.data.sql.select_by_hash(session, table_object, hash_partial)[source]

Find a record by its partial hash.

The hash provided is considered to be the first len(hash_partial) carachters of a full hash in the table.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object
hash_partial (str) – hash identifier

Returns:

record

Return type:

banana.data.sql.select_by_position(session, table_object, pos)[source]

Find a record by its position.

Negative values are supported, and used as positions from the end (exact same logic of python list).

Note

This is more expensive than using select_uid(), since the whole table is retrieved, and only after a single member is selected.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object
hash_partial (str) – hash identifier

Returns:

record

Return type:

banana.data.sql.select_by_uid(session, table_object, uid)[source]

Find a record by its uid (unique identifier).

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object
hash_partial (str) – hash identifier

Returns:

record

Return type:

banana.data.sql.serialize(data)[source]

Eventually turn some elements into their binary representation.

Parameters:: data (dict) – raw data
Returns:: ndata – improved data
Return type:: list

banana.data.sql.truncate(session, table_object)[source]

Empty table.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object

banana.data.sql.update_atime(session, table_object, uids)[source]

Update rows access time to now.

Parameters:

session (sqlalchemy.orm.session.Session) – DB ORM session
table_object (sqlalchemy.schema.Table) – table object
uids (list(int)) – unique identifiers of rows to update

banana.data.theories module

banana.data.theories.load(session, updates)[source]

Load theories.

Parameters:

session (sqlalchemy.orm.session.Session) – database
updates (list(dict)) – update directives

Returns:

list of all dictionaries

Return type: