def n_tree():
"""
This is the format of this tree
t 1
a b c 3
aa ab ba bb ca cb 6
aaa aab aba abb baa bab bba bbb caa cab cba cbb 12
Resulting in the summing matrix: y_t = S * b_t
t 1 1 1 1 1 1 1 1 1 1 1 1
a 1 1 1 1 0 0 0 0 0 0 0 0
b 0 0 0 0 1 1 1 1 0 0 0 0
c 0 0 0 0 0 0 0 0 1 1 1 1
aa 1 1 0 0 0 0 0 0 0 0 0 0
ab 0 0 1 1 0 0 0 0 0 0 0 0 aaa
ba 0 0 0 0 1 1 0 0 0 0 0 0 aab
bb 0 0 0 0 0 0 1 1 0 0 0 0 aba
ca 0 0 0 0 0 0 0 0 1 1 0 0 abb
cb 0 0 0 0 0 0 0 0 0 0 1 1 baa
aaa 1 0 0 0 0 0 0 0 0 0 0 0 bab
aab 0 1 0 0 0 0 0 0 0 0 0 0 bba
aba 0 0 1 0 0 0 0 0 0 0 0 0 bbb
abb 0 0 0 1 0 0 0 0 0 0 0 0 caa
baa 0 0 0 0 1 0 0 0 0 0 0 0 cab
bab 0 0 0 0 0 1 0 0 0 0 0 0 cba
bba 0 0 0 0 0 0 1 0 0 0 0 0 cbb
bbb 0 0 0 0 0 0 0 1 0 0 0 0
caa 0 0 0 0 0 0 0 0 1 0 0 0
cab 0 0 0 0 0 0 0 0 0 1 0 0
cba 0 0 0 0 0 0 0 0 0 0 1 0
cbb 0 0 0 0 0 0 0 0 0 0 0 1
"""
t = ('t', 1)
t1 = [('a', 2), ('b', 2), ('c', 3)]
t2 = [('aa', 4), ('ab', 5), ('ba', 6), ('bb', 4), ('ca', 5), ('cb', 6)]
t3 = [('aaa', 4), ('aab', 5), ('aba', 6), ('abb', 4), ('baa', 5),
('bab', 6), ('bba', 5), ('bbb', 6), ('caa', 5), ('cab', 6),
('cba', 5), ('cbb', 6)]
test_t = HierarchyTree(key=t[0], item=t[1])
for i, j in t1:
test_t.add_child(key=i, item=j)
for c in test_t.children:
for i, j in t2:
if i.startswith(c.key):
c.add_child(key=i, item=j)
for c in test_t.children:
for c2 in c.children:
for i, j in t3:
if i.startswith(c2.key):
c2.add_child(key=i, item=j)
return test_t
def test_create_mv_tree(hierarchical_mv_data):
hier = {
'total': ['CH', 'SLU', 'BT', 'OTHER'],
'CH': ['CH-07', 'CH-02', 'CH-08', 'CH-05', 'CH-01'],
'SLU': ['SLU-15', 'SLU-01', 'SLU-19', 'SLU-07', 'SLU-02'],
'BT': ['BT-01', 'BT-03'],
'OTHER': ['WF-01', 'CBD-13']
}
exogenous = {
k: ['precipitation', 'temp']
for k in hierarchical_mv_data.columns
if k not in ['precipitation', 'temp']
}
ht = HierarchyTree.from_nodes(hier,
hierarchical_mv_data,
exogenous=exogenous)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
assert ht.key == 'total'
assert len(ht.children) == 4
assert ht.get_node_height('CH') == 1
assert ht.get_node_height('BT-03') == 0
assert ht.get_node_height('CBD-13') == 0
assert ht.get_node_height('SLU') == 1
def __init_hts(
self,
nodes: Optional[NodesT] = None,
df: Optional[pandas.DataFrame] = None,
tree: Optional[HierarchyTree] = None,
root: str = "root",
exogenous: Optional[List[str]] = None,
):
if not nodes and not df:
if not tree:
raise InvalidArgumentException(
"Either nodes and df must be passed, or a pre-built hierarchy tree"
)
else:
self.nodes = tree
else:
self.nodes = HierarchyTree.from_nodes(nodes=nodes,
df=df,
exogenous=exogenous,
root=root)
self.exogenous = exogenous
self.sum_mat, sum_mat_labels = to_sum_mat(self.nodes)
self._set_model_instance()
self._init_revision()
def test_create_mv_tree(hierarchical_mv_data):
hier = {
"total": ["CH", "SLU", "BT", "OTHER"],
"CH": ["CH-07", "CH-02", "CH-08", "CH-05", "CH-01"],
"SLU": ["SLU-15", "SLU-01", "SLU-19", "SLU-07", "SLU-02"],
"BT": ["BT-01", "BT-03"],
"OTHER": ["WF-01", "CBD-13"],
}
exogenous = {
k: ["precipitation", "temp"]
for k in hierarchical_mv_data.columns
if k not in ["precipitation", "temp"]
}
ht = HierarchyTree.from_nodes(hier,
hierarchical_mv_data,
exogenous=exogenous)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
assert ht.key == "total"
assert len(ht.children) == 4
assert ht.get_node_height("CH") == 1
assert ht.get_node_height("BT-03") == 0
assert ht.get_node_height("CBD-13") == 0
assert ht.get_node_height("SLU") == 1
def test_to_pandas(events):
ht = HierarchyTree.from_geo_events(df=events,
lat_col='start_latitude',
lon_col='start_longitude',
nodes=('city', 'hex_index_6',
'hex_index_7', 'hex_index_8'),
levels=(6, 8),
resample_freq='1H',
min_count=0.5)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
def mv_tree(hierarchical_mv_data, mv_tree_empty):
exogenous = {
k: ["precipitation", "temp"]
for k in hierarchical_mv_data.columns
if k not in ["precipitation", "temp"]
}
return HierarchyTree.from_nodes(mv_tree_empty,
hierarchical_mv_data,
exogenous=exogenous)
def test_to_pandas(events):
ht = HierarchyTree.from_geo_events(
df=events,
lat_col="start_latitude",
lon_col="start_longitude",
nodes=("city", "hex_index_6", "hex_index_7", "hex_index_8"),
levels=(6, 8),
resample_freq="1H",
min_count=0.5,
)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
def test_create_from_events(events):
ht = HierarchyTree.from_geo_events(df=events,
lat_col='start_latitude',
lon_col='start_longitude',
nodes=('city', 'hex_index_6',
'hex_index_7', 'hex_index_8'),
levels=(6, 8),
resample_freq='1H',
min_count=0.5)
assert isinstance(ht, NAryTreeT)
assert len(ht.children) == events['city'].nunique()
def test_create_from_events(events):
ht = HierarchyTree.from_geo_events(
df=events,
lat_col="start_latitude",
lon_col="start_longitude",
nodes=("city", "hex_index_6", "hex_index_7", "hex_index_8"),
levels=(6, 8),
resample_freq="1H",
min_count=0.5,
)
assert isinstance(ht, NAryTreeT)
assert len(ht.children) == events["city"].nunique()
def mv_tree(hierarchical_mv_data):
hier = {
"total": ["CH", "SLU", "BT", "OTHER"],
"CH": ["CH-07", "CH-02", "CH-08", "CH-05", "CH-01"],
"SLU": ["SLU-15", "SLU-01", "SLU-19", "SLU-07", "SLU-02"],
"BT": ["BT-01", "BT-03"],
"OTHER": ["WF-01", "CBD-13"],
}
exogenous = {
k: ["precipitation", "temp"]
for k in hierarchical_mv_data.columns
if k not in ["precipitation", "temp"]
}
return HierarchyTree.from_nodes(hier, hierarchical_mv_data, exogenous=exogenous)
def mv_tree(hierarchical_mv_data):
hier = {
'total': ['CH', 'SLU', 'BT', 'OTHER'],
'CH': ['CH-07', 'CH-02', 'CH-08', 'CH-05', 'CH-01'],
'SLU': ['SLU-15', 'SLU-01', 'SLU-19', 'SLU-07', 'SLU-02'],
'BT': ['BT-01', 'BT-03'],
'OTHER': ['WF-01', 'CBD-13']
}
exogenous = {
k: ['precipitation', 'temp']
for k in hierarchical_mv_data.columns
if k not in ['precipitation', 'temp']
}
return HierarchyTree.from_nodes(hier,
hierarchical_mv_data,
exogenous=exogenous)
def predict(
self,
node: HierarchyTree,
freq: str = "D",
steps_ahead: int = 1,
exogenous_df: pandas.DataFrame = None,
):
df = self._pre_process(node.item)
future = self.model.make_future_dataframe(
periods=steps_ahead, freq=freq, include_history=True
)
if exogenous_df is not None:
previous_exogenous_values = node.to_pandas()[node.exogenous].reset_index(
drop=True
)
future_exogenous = pandas.concat(
[previous_exogenous_values, exogenous_df]
).reset_index(drop=True)
future = pandas.concat(
[future, future_exogenous.reindex(future.index)], axis=1
)
if self.cap:
future["cap"] = self.cap
if self.floor:
future["floor"] = self.floor
self.forecast = self.model.predict(future)
merged = pandas.merge(df, self.forecast, on="ds")
self.residual = (merged["yhat"] - merged["y"]).values
self.mse = numpy.mean(numpy.array(self.residual) ** 2)
if self.cap is not None:
self.forecast.yhat = numpy.exp(self.forecast.yhat)
self.forecast.yhat = self.transform_function.inverse_transform(
self.forecast.yhat
)
self.forecast.trend = self.transform_function.inverse_transform(
self.forecast.trend
)
for component in ["seasonal", "daily", "weekly", "yearly", "holidays"]:
if component in self.forecast.columns.tolist():
inv_transf = self.transform_function.inverse_transform(
getattr(self.forecast, component)
)
setattr(self.forecast, component, inv_transf)
return self
def test_create_hierarchical_sine_data_tree(hierarchical_sine_data):
hier = {
"total": ["a", "b", "c"],
"a": ["aa", "ab"],
"aa": ["aaa", "aab"],
"b": ["ba", "bb"],
"c": ["ca", "cb", "cc", "cd"],
}
ht = HierarchyTree.from_nodes(hier, hierarchical_sine_data)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
assert ht.key == "total"
assert len(ht.children) == 3
for c in ht.children:
if c.key == "a" or c.key == "b":
assert len(c.children) == 2
if c.key == "c":
assert len(c.children) == 4
def test_create_hierarchical_sine_data_tree(hierarchical_sine_data):
hier = {
'total': ['a', 'b', 'c'],
'a': ['aa', 'ab'],
'aa': ['aaa', 'aab'],
'b': ['ba', 'bb'],
'c': ['ca', 'cb', 'cc', 'cd']
}
ht = HierarchyTree.from_nodes(hier, hierarchical_sine_data)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
assert ht.key == 'total'
assert len(ht.children) == 3
for c in ht.children:
if c.key == 'a' or c.key == 'b':
assert len(c.children) == 2
if c.key == 'c':
assert len(c.children) == 4
def test_create_hierarchical_sine_data_tree(hierarchical_sine_data):
hier = {
"total": ["a", "b", "c"],
"a": ["a_x", "a_y"],
"b": ["b_x", "b_y"],
"c": ["c_x", "c_y"],
"a_x": ["a_x_1", "a_x_2"],
"a_y": ["a_y_1", "a_y_2"],
"b_x": ["b_x_1", "b_x_2"],
"b_y": ["b_y_1", "b_y_2"],
"c_x": ["c_x_1", "c_x_2"],
"c_y": ["c_y_1", "c_y_2"],
}
ht = HierarchyTree.from_nodes(hier, hierarchical_sine_data)
assert isinstance(ht.to_pandas(), pandas.DataFrame)
assert ht.key == "total"
assert len(ht.children) == 3
for c in ht.children:
if c.key == "a" or c.key == "b" or c.key == "c":
assert len(c.children) == 2
if (c.key == "a_x" or c.key == "b_x" or c.key == "c_x"
or c.key == "a_y" or c.key == "b_y" or c.key == "c_y"):
assert len(c.children) == 4
def uv_tree(sine_hier, hierarchical_sine_data):
hsd = hierarchical_sine_data.resample('1H').apply(sum).head(400)
return HierarchyTree.from_nodes(sine_hier, hsd)