[minor] Support data without datestamps

neuralprophet/df_utils.py

@@ -12,7 +12,6 @@
 if TYPE_CHECKING:
     from neuralprophet.configure import ConfigEvents, ConfigLaggedRegressors, ConfigSeasonality
 
-
 log = logging.getLogger("NP.df_utils")
 
 
@@ -37,11 +36,15 @@ def prep_or_copy_df(df: pd.DataFrame) -> tuple[pd.DataFrame, bool, bool, list[st
         bool
             wheter it is a single time series
         list
-            list of IDs
+            list of IDsgithubn
     """
     if not isinstance(df, pd.DataFrame):
         raise ValueError("Provided DataFrame (df) must be of pd.DataFrame type.")
-
+    if "ds" not in df and "y" in df:
+        df = create_random_datestamps(df)
+        log.info(
+            "Dataframe has no column 'ds' - random equidistant datestamps added. Consider calling 'df_utils.create_random_datestamps' to adjust ds."
+        )
     # Create a copy of the dataframe
     df_copy = df.copy(deep=True)
 
@@ -502,6 +505,33 @@ def check_single_dataframe(df, check_y, covariates, regressors, events, seasonal
     return df
 
 
+def create_random_datestamps(
+    df, freq="D", startyear=1970, startmonth=1, startday=1, starthour=0, startminute=0, startsecond=0
+):
+    """
+    Helper function to create a random series of datestamps for equidistant data without ds.
+
+    Parameters
+    ----------
+        df : pd.DataFrame
+            without column ``ds``
+        freq : str
+            Frequency of data recording, any valid frequency for pd.date_range, such as ``D`` or ``M``
+        startyear, startmonth, startday, starthour, startminute, startsecond : int
+            Defines the first datestamp
+    Returns
+    -------
+        pd.DataFrame or dict
+            dataframe with random equidistant datestamps
+    """
+    startdate = pd.Timestamp(
+        year=startyear, month=startmonth, day=startday, hour=starthour, minute=startminute, second=startsecond
+    )
+    ds = pd.date_range(startdate, periods=len(df), freq=freq).to_frame(index=False, name="ds").astype(str)
+    df_random_ds = pd.concat([ds, df], axis=1)
+    return df_random_ds
+
+
 def check_dataframe(
     df: pd.DataFrame, check_y: bool = True, covariates=None, regressors=None, events=None, seasonalities=None
 ) -> Tuple[pd.DataFrame, List]:

tests/test_utils.py

@@ -26,6 +26,17 @@
 PLOT = False
 
 
+def test_create_random_datestamps():
+    df = pd.read_csv(PEYTON_FILE)
+    df_drop = df.drop("ds", axis=1)
+
+    m = NeuralProphet(quantiles=[0.02, 0.98], epochs=10)
+    _ = m.fit(df_drop, freq="D")
+
+    future = m.make_future_dataframe(df_drop, periods=365, n_historic_predictions=True)
+    forecast = m.predict(future)
+
+
 def test_save_load():
     df = pd.read_csv(PEYTON_FILE, nrows=NROWS)
     m = NeuralProphet(