I am using the prophet forecaster from python’s sktime package

https://www.sktime.net/en/stable/api_reference/auto_generated/sktime.forecasting.fbprophet.Prophet.html

I am not sure how to add custom seasonalities correctly and how to consider them in the fit. I tried the following:

import pandas as pd
from sktime.forecasting.fbprophet import Prophet

# Example data creation
dates = pd.date_range(start='2023-01-01', end='2023-01-31')
data = {'dt': dates, 'y': range(len(dates))}
df_train = pd.DataFrame(data)

# Create a custom 'weekday_flag' column
df_train['weekday_flag'] = df_train['dt'].dt.weekday < 5  # Monday=0, ..., Friday=4

# Define custom seasonality with condition 'is_weekday'
custom_seasonality = [
    {
        "name": "weekday_seasonality",
        "period": 1, 
        "fourier_order": 3,
        "condition_name": "weekday_flag",
    }
    ]

# Initialize Prophet model with custom seasonality
prophet_model = Prophet(add_seasonality=custom_seasonality)

# Fit the Prophet model using the features and target
X = df_train[['weekday_flag']].set_index(df_train['dt'])
y = df_train[['y']].set_index(df_train['dt'])

prophet_model.fit(X=X, y=y)

But I am getting the error TypeError: ufunc 'invert' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''.

Any hints how to resolve it? Do I need to change the format of X and y?