导入JSON数据后遇到问题。通常_to_numeric_可以正常工作,但在这种情况下则不能。我有一串字符串看起来像这样: 10.2100 10.2400 ...
我正在使用Jupyter Notebook。 我已经尝试过 astype(),_ infer_objects(),但是数据仍然表现得像字符串一样。我通过将列乘以2来检查并得到: 10.210010.2100 10.240010.2400 ...
这是代码:
df = pd.DataFrame()
temp = pd.DataFrame()
for file in filelist:
with open(path+file, "r", encoding="windows-1252") as f:
xml_string = f.read()
if xml_string.find("<") >= 0:
json_string = convert_to_json(xml_string)
flat = flatten_json.flatten(json.loads(json_string))
temp = pd.DataFrame.from_dict(flat, orient='index').transpose()
df = df.append(temp, sort=False)
df = df.reset_index()
pd.to_numeric(df.['Telegram_PRC_VALS_STP_2_VAL_0_NUM'])
df['Telegram_PRC_VALS_STP_2_VAL_0_NUM'] = df['Telegram_PRC_VALS_STP_2_VAL_0_NUM']*2
´´´
Any ideas? As this is my first Question in here I hope I am meeting your expectations on decently asking Questions.
答案 0 :(得分:2)
您必须重新分配to_numeric
的输出,因为它不是就地方法:
df['Telegram_PRC_VALS_STP_2_VAL_0_NUM']=pd.to_numeric(df['Telegram_PRC_VALS_STP_2_VAL_0_NUM'])