Plots in forecast tutorial doesn't match output.

[BinhHong]

I run the code in Prophet Forecast in tutorial with the air_passengers dataset and m.plot() but the result doesn't match the output there. If I remember correctly, it used to.

[MoKazemi9]

Hi @BinhHong ,Thanks for sharing the issue!, can you provide more context, the code and the plot to replicate the issue?

[BinhHong]

Hi @MoKazemi9 , the code here is exactly the same as in tutorial https://github.com/facebookresearch/Kats/blob/main/tutorials/kats_201_forecasting.ipynb.

The dataset is air_passengers.csv

air_passengers_df = pd.read_csv("air_passengers.csv")
air_passengers_df.columns = ["time", "value"]
air_passengers_ts = TimeSeriesData(air_passengers_df)

from kats.models.sarima import SARIMAModel, SARIMAParams
warnings.simplefilter(action='ignore')

# create SARIMA param class
params = SARIMAParams(
    p = 2, 
    d=1, 
    q=1, 
    trend = 'ct', 
    seasonal_order=(1,0,1,12)
    )

# initiate SARIMA model
m = SARIMAModel(data=air_passengers_ts, params=params)

# fit SARIMA model
m.fit()

# generate forecast values
fcst = m.predict(
    steps=30, 
    freq="MS"
    )

Then m.plot() produces whereas in the tutorial

[SakshiNYCU]

@BinhHong I'm facing the same issue with the Ensemble model. A couple of weeks ago my graphs showed the actual and forecasted trends together but now as I try to replicate my results with the same code and data, they just show the forecasted trends.

I suggest you try the following:

# generate forecast values fcst = m.predict( steps=30, freq="MS" include_history = True )

I tried it for Ensemble, but it shows some error. However, in the case of SARIMA it works for me.

[BinhHong]

Hi @SakshiNYCU, thanks a lot for your suggestion. The result looks as I expected

compared to the tutorial

[jeffhandl]

263 should resolve this.