当我尝试使用delta.io库在AWS Glue中创建DeltaTable时，我收到此错误：

Execution error : field item_size: ArrayType(DoubleType,true) can not accept object array([19. ,  5. ,  6.5]) in type <class 'numpy.ndarray'>

下面是我的一些代码：

df = wr.athena.read_sql_query(sql=query, database="db_name")
    registration_schema = --more code--
    StructField("item_size",DoubleType(),nullable=True)
    --more code--

    df_delta = spark.createDataFrame(df,schema=registration_schema)

df包含：

item_id          item_size  ...             acquiredate      acquiredate_tz
74        3041    [7.0, 5.0, 5.0]  ... 2022-10-25 10:30:15.974 2022-10-25 12:30:15
152       3142   [19.0, 7.0, 5.0]  ... 2022-10-25 10:29:47.985 2022-10-25 12:29:47
154       2696  [31.0, 2.5, 10.0]  ... 2022-10-25 10:29:50.838 2022-10-25 12:29:50
158       2198   [22.1, 6.1, 6.1]  ... 2022-10-25 10:29:54.353 2022-10-25 12:29:54
251       2593  [4.0, 15.0, 20.0]  ... 2022-10-25 10:29:51.636 2022-10-25 12:29:51

df.info(verbose=True)
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 12069 entries, 0 to 12068
Data columns (total 12 columns):
 #   Column               Non-Null Count  Dtype         
---  ------               --------------  -----         
 0   item_id              12068 non-null  Int64         
 1   item_size            324 non-null    object        
 2   box_size             178 non-null    object

我该如何解决这个问题？