TimeSeriesDataFrame.from_path#

classmethod TimeSeriesDataFrame.from_path(path: Union[str, Path], id_column: Optional[str] = None, timestamp_column: Optional[str] = None, static_features_path: Optional[Union[str, Path]] = None) → TimeSeriesDataFrame[source]#

Construct a TimeSeriesDataFrame from a CSV or Parquet file.

Parameters

path (str or pathlib.Path) –

Path to a local or remote (e.g., S3) file containing the time series data in CSV or Parquet format. Example file contents:

item_id,timestamp,target
0,2019-01-01,0
0,2019-01-02,1
0,2019-01-03,2
1,2019-01-01,3
1,2019-01-02,4
1,2019-01-03,5
2,2019-01-01,6
2,2019-01-02,7
2,2019-01-03,8

id_column (str, optional) – Name of the ‘item_id’ column if column name is different
timestamp_column (str, optional) – Name of the ‘timestamp’ column if column name is different
static_features_path (str or pathlib.Path, optional) –
Path to a local or remote (e.g., S3) file containing static features in CSV or Parquet format. Example file contents:
```
item_id,feat_1,feat_2
0,foo,0.5
1,foo,2.2
2,bar,0.1
```

Returns

ts_df – A data frame in TimeSeriesDataFrame format.

Return type

TimeSeriesDataFrame