Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fixes date and frequency issues in forecasting #1094

Merged
merged 2 commits into from
Sep 12, 2023
Merged

Fixes date and frequency issues in forecasting #1094

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5b832a5
fix date and frequency issues
americast Sep 11, 2023
89cca43
Update docs
americast Sep 12, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 8 additions & 5 deletions evadb/executor/create_function_executor.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -149,15 +149,11 @@ def handle_forecasting_function(self):
impl_path = Path(f"{self.function_dir}/forecast.py").absolute().as_posix()
else:
impl_path = self.node.impl_path.absolute().as_posix()
arg_map = {arg.key: arg.value for arg in self.node.metadata}

if "model" not in arg_map.keys():
arg_map["model"] = "AutoARIMA"
if "frequency" not in arg_map.keys():
arg_map["frequency"] = "M"

model_name = arg_map["model"]
frequency = arg_map["frequency"]

"""
The following rename is needed for statsforecast, which requires the column name to be the following:
Expand All @@ -179,6 +175,10 @@ def handle_forecasting_function(self):
if "ds" not in list(data.columns):
data["ds"] = [x + 1 for x in range(len(data))]

if "frequency" not in arg_map.keys():
arg_map["frequency"] = pd.infer_freq(data["ds"])
frequency = arg_map["frequency"]

try_to_import_forecast()
from statsforecast import StatsForecast
from statsforecast.models import AutoARIMA, AutoCES, AutoETS, AutoTheta
Expand Down Expand Up @@ -220,7 +220,7 @@ def handle_forecasting_function(self):
)

weight_file = Path(model_path)

data["ds"] = pd.to_datetime(data["ds"])
if not weight_file.exists():
model.fit(data)
f = open(model_path, "wb")
Expand All @@ -233,6 +233,9 @@ def handle_forecasting_function(self):
FunctionMetadataCatalogEntry("model_name", model_name),
FunctionMetadataCatalogEntry("model_path", model_path),
FunctionMetadataCatalogEntry("output_column_rename", arg_map["predict"]),
FunctionMetadataCatalogEntry(
"time_column_rename", arg_map["time"] if "time" in arg_map else "ds"
),
]

return (
Expand Down
14 changes: 12 additions & 2 deletions evadb/functions/forecast.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,13 +28,20 @@ def name(self) -> str:
return "ForecastModel"

@setup(cacheable=False, function_type="Forecasting", batchable=True)
def setup(self, model_name: str, model_path: str, output_column_rename: str):
def setup(
self,
model_name: str,
model_path: str,
output_column_rename: str,
time_column_rename: str,
):
f = open(model_path, "rb")
loaded_model = pickle.load(f)
f.close()
self.model = loaded_model
self.model_name = model_name
self.output_column_rename = output_column_rename
self.time_column_rename = time_column_rename

def forward(self, data) -> pd.DataFrame:
horizon = list(data.iloc[:, -1])[0]
Expand All @@ -43,6 +50,9 @@ def forward(self, data) -> pd.DataFrame:
), "Forecast UDF expects integral horizon in parameter."
forecast_df = self.model.predict(h=horizon)
forecast_df = forecast_df.rename(
columns={self.model_name: self.output_column_rename}
columns={
self.model_name: self.output_column_rename,
"ds": self.time_column_rename,
}
)
return forecast_df