无法摆脱numpy数组中的非Unicode字符

Question

I have a table: vals = table.iloc[sum(which(pa_i),[1]),1:] content:

Out[249]: 
                      1         2       3
2  NA [1] (16.0 to N/A)  12.0 [2]  NA [1]

When I write vals.values I get: vals

Out[250]: array([['NA\xa0[1] (16.0\xa0to\xa0N/A)', '12.0\xa0[2]', 'NA\xa0[1]']], dtype=object)

I simply want to get this table to an array or list but earlier want to replace \\xa0 values. When I apply: v = print(map(lambda s: s.replace('\\xa0' , ' '), vals)) I don't know how to read this object or when np.char.replace(vals[0],'\\xa0' , ' ') I get error "TypeError: string operation on non-string array".

What is the easiest way to convert the content to array or to replace unwanted chars??

EDIT

I've got a solution: v = vals.astype('str') and v = np.char.replace(v,'\\xa0' , ' ') .

Out[306]: array([['NA [1] (16.0 to N/A)', '12.0 [2]', 'NA [1]']], dtype='<U20') 

But I'm not fully satisfied of this answer. I need something to work directly on vals variable - for example by doing this: a = vals[1:].toarray(?) expected result:

a
Out[318]: ['(16.0 to N/A)', '12.0 [2]', 'NA [1]]']

Answer 1

Well, its because it contains int elements and not str

Do this:

table.values = [str(v) for v in table.values]

Answer 2

I have found the best way:

vals
Out[357]: 
                      1         2       3
2  NA [1] (16.0 to N/A)  12.0 [2]  NA [1]

a = vals.values.tolist()

a
Out[358]: [['NA\xa0[1] (16.0\xa0to\xa0N/A)', '12.0\xa0[2]', 'NA\xa0[1]']]

a1 = [w.replace('\\xa0', ' ') for w in a[0]]

a1
Out[359]: ['NA [1] (16.0 to N/A)', '12.0 [2]', 'NA [1]']

Pffff. Maybe not the prettiest in terms of coding habits but what can we do.