# Copyright 2022 The SQLNet Company GmbH
#
# This file is licensed under the Elastic License 2.0 (ELv2).
# Refer to the LICENSE.txt file in the root of the repository
# for details.
#
"""
Contains a wrapper around connect.
"""
import sqlite3
from .contains import _contains
from .count_above_mean import _CountAboveMean
from .count_below_mean import _CountBelowMean
from .count_distinct_over_count import _CountDistinctOverCount
from .email_domain import _email_domain
from .ewma import (
_EWMA1S,
_EWMA1M,
_EWMA1H,
_EWMA1D,
_EWMA7D,
_EWMA30D,
_EWMA90D,
_EWMA365D,
)
from .ewma_trend import (
_EWMATrend1S,
_EWMATrend1M,
_EWMATrend1H,
_EWMATrend1D,
_EWMATrend7D,
_EWMATrend30D,
_EWMATrend90D,
_EWMATrend365D,
)
from .first import _First
from .get_word import _get_word
from .kurtosis import _Kurtosis
from .last import _Last
from .median import _Median
from .mode import _Mode
from .num_max import _NumMax
from .num_min import _NumMin
from .num_words import _num_words
from .quantiles import (
_Q1,
_Q5,
_Q10,
_Q25,
_Q75,
_Q90,
_Q95,
_Q99,
)
from .skew import _Skew
from .stddev import _Stddev
from .time_since_first_maximum import _TimeSinceFirstMaximum
from .time_since_first_minimum import _TimeSinceFirstMinimum
from .time_since_last_maximum import _TimeSinceLastMaximum
from .time_since_last_minimum import _TimeSinceLastMinimum
from .trend import _Trend
from .var import _Var
from .variation_coefficient import _VariationCoefficient
[docs]def connect(database: str):
"""
Generates a new sqlite3 connection.
This connection contains all customized aggregations
and transformation functions needed to execute the
SQL pipeline generated by getML. Other than that
it behaves just like a normal sqlite3 connection from
the Python standard library.
Args:
database (str):
Filename of the database. Use ':memory:' to
create an in-memory database.
"""
if not isinstance(database, str):
raise TypeError("'database' must be of type str")
if sqlite3.sqlite_version < "3.33.0":
raise ValueError(
"getML requires SQLite version 3.33.0 or above. Found version "
+ sqlite3.sqlite_version
+ ". Please upgrade Python and/or the Python sqlite3 package."
)
conn = sqlite3.connect(database)
conn.create_function("contains", 2, _contains)
conn.create_function("email_domain", 1, _email_domain)
conn.create_function("get_word", 2, _get_word)
conn.create_function("num_words", 1, _num_words)
conn.create_aggregate("COUNT_ABOVE_MEAN", 1, _CountAboveMean) # type: ignore
conn.create_aggregate("COUNT_BELOW_MEAN", 1, _CountBelowMean) # type: ignore
conn.create_aggregate("COUNT_DISTINCT_OVER_COUNT", 1, _CountDistinctOverCount) # type: ignore
conn.create_aggregate("EWMA_1S", 2, _EWMA1S) # type: ignore
conn.create_aggregate("EWMA_1M", 2, _EWMA1M) # type: ignore
conn.create_aggregate("EWMA_1H", 2, _EWMA1H) # type: ignore
conn.create_aggregate("EWMA_1D", 2, _EWMA1D) # type: ignore
conn.create_aggregate("EWMA_7D", 2, _EWMA7D) # type: ignore
conn.create_aggregate("EWMA_30D", 2, _EWMA30D) # type: ignore
conn.create_aggregate("EWMA_90D", 2, _EWMA90D) # type: ignore
conn.create_aggregate("EWMA_365D", 2, _EWMA365D) # type: ignore
conn.create_aggregate("EWMA_TREND_1S", 2, _EWMATrend1S) # type: ignore
conn.create_aggregate("EWMA_TREND_1M", 2, _EWMATrend1M) # type: ignore
conn.create_aggregate("EWMA_TREND_1H", 2, _EWMATrend1H) # type: ignore
conn.create_aggregate("EWMA_TREND_1D", 2, _EWMATrend1D) # type: ignore
conn.create_aggregate("EWMA_TREND_7D", 2, _EWMATrend7D) # type: ignore
conn.create_aggregate("EWMA_TREND_30D", 2, _EWMATrend30D) # type: ignore
conn.create_aggregate("EWMA_TREND_90D", 2, _EWMATrend90D) # type: ignore
conn.create_aggregate("EWMA_TREND_365D", 2, _EWMATrend365D) # type: ignore
conn.create_aggregate("FIRST", 2, _First) # type: ignore
conn.create_aggregate("KURTOSIS", 1, _Kurtosis)
conn.create_aggregate("LAST", 2, _Last) # type: ignore
conn.create_aggregate("MEDIAN", 1, _Median)
conn.create_aggregate("MODE", 1, _Mode)
conn.create_aggregate("NUM_MAX", 1, _NumMax) # type: ignore
conn.create_aggregate("NUM_MIN", 1, _NumMin) # type: ignore
conn.create_aggregate("Q1", 1, _Q1) # type: ignore
conn.create_aggregate("Q5", 1, _Q5) # type: ignore
conn.create_aggregate("Q10", 1, _Q10) # type: ignore
conn.create_aggregate("Q25", 1, _Q25) # type: ignore
conn.create_aggregate("Q75", 1, _Q75) # type: ignore
conn.create_aggregate("Q90", 1, _Q90) # type: ignore
conn.create_aggregate("Q95", 1, _Q95) # type: ignore
conn.create_aggregate("Q99", 1, _Q99) # type: ignore
conn.create_aggregate("SKEW", 1, _Skew)
conn.create_aggregate("STDDEV", 1, _Stddev)
conn.create_aggregate("TIME_SINCE_FIRST_MAXIMUM", 2, _TimeSinceFirstMaximum) # type: ignore
conn.create_aggregate("TIME_SINCE_FIRST_MINIMUM", 2, _TimeSinceFirstMinimum) # type: ignore
conn.create_aggregate("TIME_SINCE_LAST_MAXIMUM", 2, _TimeSinceLastMaximum) # type: ignore
conn.create_aggregate("TIME_SINCE_LAST_MINIMUM", 2, _TimeSinceLastMinimum) # type: ignore
conn.create_aggregate("TREND", 2, _Trend) # type: ignore
conn.create_aggregate("VAR", 1, _Var)
conn.create_aggregate("VARIATION_COEFFICIENT", 1, _VariationCoefficient)
return conn