Python: endog has evaluated to an array with multiple columns

Question

I'm trying to run a Poisson model, like this:

poisson_model_xg = smf.glm(formula="xG ~ home + team + opponent", data=xg_model_data, 
                        family=sm.families.Poisson()).fit()

I'm getting the following error:

ValueError: endog has evaluated to an array with multiple columns that has shape (760, 9). This occurs when the variable converted to endog is non-numeric (e.g., bool or str).

But I can't figure out what does it mean, since all my dataframe is numeric:

xg_model_data.apply(lambda s: pd.to_numeric(s, errors='coerce').notnull().all())
Out[10]: 
goals       True
xG          True
team        True
opponent    True
home        True
dtype: bool

Sussna

Answer 1

Solved. The trick was not in content type, but in columns type:

xg_model_data.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 760 entries, 0 to 759
Data columns (total 5 columns):
 #   Column    Non-Null Count  Dtype 
---  ------    --------------  ----- 
 0   goals     760 non-null    object
 1   xG        760 non-null    object
 2   team      760 non-null    object
 3   opponent  760 non-null    object
 4   home      760 non-null    object
dtypes: object(5)
memory usage: 55.6+ KB

After I applied pd.to_numeric() on desired columns, the dataframe looks like the following, and Poisson is able to process.

xg_model_data.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 760 entries, 0 to 759
Data columns (total 5 columns):
 #   Column    Non-Null Count  Dtype  
---  ------    --------------  -----  
 0   goals     760 non-null    int64  
 1   xG        760 non-null    float64
 2   team      760 non-null    object 
 3   opponent  760 non-null    object 
 4   home      760 non-null    int64  
dtypes: float64(1), int64(2), object(2)
memory usage: 55.6+ KB

Sussna

and "applying" means df['col_name'] = pd.to_numeric(df['col_name']) – Histo