python - how can i define dtypes when using io.parsers.read_csv -

When I load my data with the following code using Iio.parser.read_csv:

< NP name_header = ['Date', 'A', 'B', 'C', 'D', 'D', 'F', '' '' ' G ',' H] df_csv_data = pd .io.parsers.read_csv (station_path, skiprows = 5, parse_dates = true, index_col = 0, na_values ​​= [- 999], names = names_header, sep = ';', header = None, Squeeze = true)

I have found this dot: but one should also have 64 flats - how do I convert it Can I have /?

  & gt; & Gt; & Gt; Df_csv_data.dtypes An Object B float 64C Float 64D Float 64E Float 64F Float 64G Float 64H Float 64 Dtepe: Object When I define it to define a dot I am using:  
  dtype = {'a': np.float64}  
 
 I have received this error message: 
 
  ValueError: can not be safely altered  
 
 The object has passed DTP of dtyped data for the object