def plot(self):
"""
---------------------------------------------------------------------------
Draws the model is the number of predictors is 2 or 3.
"""
if (2 <= len(self.X) <= 3):
vDataFrame(self.name, self.cursor).scatter(columns=self.X,
catcol="dbscan_cluster",
max_cardinality=100,
max_nb_points=10000)
else:
raise ValueError("Clustering Plots are only available in 2D or 3D")
def deploy_to_DB(self, name: str, view: bool = True, cutoff: float = -1):
"""
---------------------------------------------------------------------------
Deploys the model in the Vertica DB by creating a relation.
Parameters
----------
name: str
Relation name. It must include the schema (the default schema is public).
view: bool
If set to false, it will create a table instead of a view.
cutoff: float, optional
Probability cutoff. If this number is not between 0 and 1, a column
corresponding to the probability to be of class 1 will be generated.
Returns
-------
vDataFrame
the vDataFrame of the new relation.
"""
check_types([("name", name, [str], False),
("view", view, [bool], False),
("cutoff", cutoff, [int, float], False)])
relation = "TABLE" if not (view) else "VIEW"
sql = "CREATE {} {} AS SELECT {}, {} AS {} FROM {}".format(
relation, name, ", ".join(self.X), self.deploySQL(cutoff), self.y,
self.test_relation)
self.cursor.execute(sql)
return vDataFrame(name, self.cursor)
def plot(self, voronoi: bool = False):
"""
---------------------------------------------------------------------------
Draws the KMeans clusters.
Parameters
----------
voronoi: bool, optional
If set to true, a voronoi plot will be drawn. It is only available for
KMeans using 2 predictors.
"""
if (voronoi):
if (len(self.X) == 2):
from vertica_ml_python.learn.plot import voronoi_plot
query = "SELECT GET_MODEL_ATTRIBUTE(USING PARAMETERS model_name = '{}', attr_name = 'centers')".format(
self.name)
self.cursor.execute(query)
clusters = self.cursor.fetchall()
voronoi_plot(clusters=clusters, columns=self.X)
else:
raise ValueError("Voronoi Plots are only available in 2D")
else:
vdf = vDataFrame(self.input_relation, self.cursor)
self.predict(vdf, "kmeans_cluster")
if (len(self.X) <= 3):
vdf.scatter(columns=self.X,
catcol="kmeans_cluster",
max_cardinality=100,
max_nb_points=10000)
else:
raise ValueError(
"Clustering Plots are only available in 2D or 3D")
def to_vdf(self): """ --------------------------------------------------------------------------- Creates a vDataFrame of the model. Returns ------- vDataFrame model vDataFrame """ return (vDataFrame(self.name, self.cursor))
def Balance(name: str,
input_relation: str,
y: str,
cursor=None,
method: str = "hybrid",
ratio: float = 0.5):
"""
---------------------------------------------------------------------------
Creates a view with an equal distribution of the input data based on the
response_column.
Parameters
----------
name: str
Name of the the view.
input_relation: str
Relation used to create the new relation.
y: str
Response column.
cursor: DBcursor, optional
Vertica DB cursor.
method: str, optional
Method used to do the balancing.
hybrid : Performs over-sampling and under-sampling on different
classes so each class is equally represented.
over : Over-samples on all classes, with the exception of the
most majority class, towards the most majority class's cardinality.
under : Under-samples on all classes, with the exception of the most
minority class, towards the most minority class's cardinality.
ratio: float, optional
The desired ratio between the majority class and the minority class. This
value has no effect when used with balance method 'hybrid'.
Returns
-------
vDataFrame
vDataFrame of the created view
"""
check_types([("name", name, [str], False),
("input_relation", input_relation, [str], False),
("y", y, [str], False),
("method", method, ["hybrid", "over", "under"], True),
("ratio", ratio, [float], False)])
if not (cursor):
cursor = read_auto_connect().cursor()
else:
check_cursor(cursor)
method = method.lower()
sql = "SELECT BALANCE('{}', '{}', '{}', '{}_sampling' USING PARAMETERS sampling_ratio = {})".format(
name, input_relation, y, method, ratio)
cursor.execute(sql)
return (vDataFrame(name, cursor))
def load_amazon(cursor=None, schema: str = 'public', name: str = 'amazon'):
"""
---------------------------------------------------------------------------
Ingests the amazon dataset in the Vertica DB (Dataset ideal for TS and
Regression). If a table with the same name and schema already exists,
this function will create a vDataFrame from the input relation.
Parameters
----------
cursor: DBcursor, optional
Vertica DB cursor.
schema: str, optional
Schema of the new relation. The default schema is public.
name: str, optional
Name of the new relation.
Returns
-------
vDataFrame
the amazon vDataFrame.
See Also
--------
load_iris : Ingests the iris dataset in the Vertica DB.
(Clustering / Classification).
load_market : Ingests the market dataset in the Vertica DB.
(Basic Data Exploration).
load_smart_meters : Ingests the smart meters dataset in the Vertica DB.
(Time Series / Regression).
load_titanic : Ingests the titanic dataset in the Vertica DB.
(Classification).
load_winequality : Ingests the winequality dataset in the Vertica DB.
(Regression / Classification).
"""
check_types([("schema", schema, [str], False),
("name", name, [str], False)])
if not (cursor):
cursor = read_auto_connect().cursor()
else:
check_cursor(cursor)
try:
vdf = vDataFrame(name, cursor, schema=schema)
except:
cursor.execute(
"CREATE TABLE {}.{}(\"number\" Integer, \"date\" Date, \"state\" Varchar(32));"
.format(str_column(schema), str_column(name)))
try:
path = os.path.dirname(
vertica_ml_python.__file__) + "/learn/data/amazon.csv"
query = "COPY {}.{}(\"number\", \"date\", \"state\") FROM {} DELIMITER ',' NULL '' ENCLOSED BY '\"' ESCAPE AS '\\' SKIP 1;".format(
str_column(schema), str_column(name), "{}")
if ("vertica_python" in str(type(cursor))):
with open(path, "r") as fs:
cursor.copy(query.format('STDIN'), fs)
else:
cursor.execute(query.format("LOCAL '{}'".format(path)))
vdf = vDataFrame(name, cursor, schema=schema)
except:
cursor.execute("DROP TABLE {}.{}".format(str_column(schema),
str_column(name)))
raise
return (vdf)