Prophet Optimization error about prophet

[Borgna02]

hi Borgna02. I had the same problem like you about the

self.stan_fit = self.model.optimize(**args)

I want to ask you how to handle this error if you already solve it

Hi, i solved it executing the script on ubuntu 22.04, prophet 1.1.1 and holidays 0.18.0.

I don't know if at the moment is possible to execute it on windows..

[2victor2]

I'm also with the same problem... digging deeper into the logs I saw that this error is being raised in cmdstanpy backend.

RuntimeError: Error during optimization! Command '/home/victor/Projects/Infra/natto-forecaster/venv/lib/python3.9/site-packages/prophet/stan_model/prophet_model.bin random seed=99561 data file=/tmp/tmpdy1bc6ti/l6ubkhnj.json init=/tmp/tmpdy1bc6ti/5gzpegdq.json output file=/tmp/tmpdy1bc6ti/prophet_modelxvyk50xw/prophet_model-20231201112246.csv method=optimize algorithm=newton iter=10000' failed: console log output:

method = optimize
  optimize
    algorithm = newton
      newton
    jacobian = 0 (Default)
    iter = 10000
    save_iterations = 0 (Default)
id = 1 (Default)
data
  file = /tmp/tmpdy1bc6ti/l6ubkhnj.json
init = /tmp/tmpdy1bc6ti/5gzpegdq.json
random
  seed = 99561
output
  file = /tmp/tmpdy1bc6ti/prophet_modelxvyk50xw/prophet_model-20231201112246.csv
  diagnostic_file =  (Default)
  refresh = 100 (Default)
  sig_figs = -1 (Default)
  profile_file = profile.csv (Default)
num_threads = 1 (Default)

Rejecting initial value:
  Error evaluating the log probability at the initial value.
Exception: normal_lpdf: Random variable is nan, but must be not nan! (in 'prophet.stan', line 130, column 2 to column 19)

Initialization between (-2, 2) failed after 1 attempts. 
 Try specifying initial values, reducing ranges of constrained values, or reparameterizing the model.
Initialization failed.

I've already tried changing the parameters, specifying initial values and clearing Nan values but, since doesn't seem to be in Prophet initialization, any of them worked. My code if needed:

        # Create a new Prophet model with the loaded parameters
        prophet_model = Prophet(**model)

        # Fit the model with the data
        df_for_prophet = data[["DTNEG", "QTDNEG"]].rename(
            columns={"DTNEG": "ds", "QTDNEG": "y"}
        )
        # Convert 'y' column to numeric, coercing errors
        df_for_prophet["y"] = pd.to_numeric(df_for_prophet["y"], errors="coerce")

        # Drop rows with NaN values after conversion
        df_for_prophet = df_for_prophet.dropna(subset=["y"])
        # Set options to display more rows and columns
        pd.set_option("display.max_rows", None)
        pd.set_option("display.max_columns", None)
        # print(
        #     {
        #         "pickle_parameters": {**model},
        #         "Loaded Prophet model for product": product,
        #         "model_name": model_name,
        #         "df": df_for_prophet,
        #     }
        # )
        prophet_model.fit(df_for_prophet) # error is being raised inside this function
        # Reset display options to default values
        pd.reset_option("display.max_rows")
        pd.reset_option("display.max_columns")

Thanks in advance

[Borgna02]

Same problem, prophet 1.1.5:

16:35:26 - cmdstanpy - INFO - Chain [1] start processing
16:35:26 - cmdstanpy - INFO - Chain [1] done processing
16:35:26 - cmdstanpy - ERROR - Chain [1] error: terminated by signal 3221225657
Optimization terminated abnormally. Falling back to Newton.
16:35:27 - cmdstanpy - INFO - Chain [1] start processing
16:35:27 - cmdstanpy - INFO - Chain [1] done processing
16:35:27 - cmdstanpy - ERROR - Chain [1] error: terminated by signal 3221225657
Traceback (most recent call last):
File "C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\prophet\models.py", line 121, in fit
self.stan_fit = self.model.optimize(**args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\cmdstanpy\model.py", line 659, in optimize
raise RuntimeError(msg)
RuntimeError: Error during optimization! Command 'C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\prophet\stan_model\prophet_model.bin random seed=82529 data file=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\chl1mqe7.json init=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\cfjxi_pi.json output file=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\prophet_modelq8hb3fxt\prophet_model-20240421163526.csv method=optimize algorithm=lbfgs iter=10000' failed:

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "d:\Uni\3 anno\1 sem\AI\progetti_corso_AI\progetto_finale\src\fbprophet\fbprhphet_prova.py", line 73, in
model.fit(future_df.reset_index().rename(columns={'index': 'ds', 'daily_return_NVDA': 'y'}))
File "C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\prophet\forecaster.py", line 1232, in fit
self.params = self.stan_backend.fit(stan_init, dat, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\prophet\models.py", line 128, in fit
self.stan_fit = self.model.optimize(**args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\cmdstanpy\model.py", line 659, in optimize
raise RuntimeError(msg)
RuntimeError: Error during optimization! Command 'C:\Users\danyb\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\LocalCache\local-packages\Python311\site-packages\prophet\stan_model\prophet_model.bin random seed=9043 data file=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\s2v4x2tf.json init=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\8u4b2w1y.json output file=C:\Users\danyb\AppData\Local\Temp\tmpz_z9dcxk\prophet_modelqfm1twda\prophet_model-20240421163527.csv method=optimize algorithm=newton iter=10000' failed:

[Tuyen92]

hi Borgna02. I had the same problem like you about the

self.stan_fit = self.model.optimize(**args)

I want to ask you how to handle this error if you already solve it