（rpy2）pandas2ri.ri2py（df）将NA_integer_转换为整数1

Question

我正在尝试使用以下代码通过rpy2将SPSS文件导入python。

from rpy2.robjects import pandas2ri, r
filename = 'x.sav'
df = r('foreign::read.spss("%s", reencode = "cp1252", use.value.labels=TRUE, to.data.frame = TRUE)' % filename)

我的文件中缺少要保留的值。 例如：

df[361]

#Returns
R object with classes: ('factor',) mapped to:
<FactorVector - Python:0x0000000011657CC8 / R:0x0000000012DAE738>
[NA_integer_, NA_integer_, NA_integer_, NA_integer_, ..., NA_integer_, NA_integer_, NA_integer_, NA_integer_]

print(df[361])

#Returns
[1] <NA>              <NA>              <NA>              <NA>             

[2] <NA>              <NA>              <NA>              <NA>             

[3] <NA>              <NA>              <NA>              <NA>                        

[4] Agree (2)         Agree (2)         <NA>              <NA>             

[5] <NA>              <NA>              <NA>              <NA>             

5 Levels: Strongly Agree (1) Agree (2) ... Strongly Disagree (5)

但是，当我执行pandas2ri.ri2py转换时，它将所有NA_interger_更改为整数值1，并接在值标签后面。

df = pandas2ri.ri2py(df)
df.iloc[:, 361]

#Returns
1      Strongly Agree (1)
2      Strongly Agree (1)
3      Strongly Agree (1)
4      Strongly Agree (1)
...
13     Agree (2)
14     Agree (2)
...
19     Strongly Agree (1)
20     Strongly Agree (1)

如何停止阻止pandas2ri.ri2py（）自动转换这些NA_Integers_？