Write a function that loops through the columns in the dataframe and applies try .astype('float64') to them. If a string value is encountered, it will be assigned the value NaN.

def convert_float(dataframe):
  for c in dataframe.columns:
    try:
      dataframe[c] = dataframe[c].astype('float64')
    except:
      pass
    
    
    
convert_float(dataframe)