如何使用.loc将新列添加到panda数据帧中

Question

我正在数据框架中计算非常简单的每日库存计算(例如SMA、VWAP、RSI等(。在我升级到anaconda3.0后，我的代码停止工作，并给出了以下错误。我在编码方面没有太多经验，需要一些帮助。

KeyError："；不再支持将缺少标签的列表likes传递给.loc或[]。缺少以下标签：索引(['RSI'，'ZONE']，dtype='object'(。看见https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#deprecate-loc reindex listlike"；

下面是代码。

import yfinance as yf
import pandas as pd
def convert_to_dataframe_daily(data):

window = 10
window20 = 20
window50 = 50
window100 = 100
window200 = 200
ema_time = 8
#data = yf.download("googl", period="30d", interval="5m")
#data = yf.download('TSLA', period='30d', interval='5m')

pd.set_option('display.max_columns', None)
#calculation for VWAP
volumeC = data['Volume']
priceC = data['Close']
df = data.assign(VWAP=((volumeC * priceC).cumsum() / volumeC.cumsum()).ffill())
#Convert the timezone to Chicago central
#df.index = pd.DatetimeIndex(df.index.tz_convert('US/Central')) # aware--> aware
#reset the dataframe index and separate time
df.reset_index(inplace=True)
#df.index.intersection
#df2 = df[df.index.isin(dts)]
#df['Date'] = pd.to_datetime(df['Datetime']).dt.date
#df['Time'] = pd.to_datetime(df['Datetime']).dt.time

# calculate stochastic

df['low5']= df['Low'].rolling(5).min()
df['high5']= df['High'].rolling(5).max()

#k = 100 * (c - l) / (h - l) 
df['K'] = (df['Close']-df['low5'])/(df['high5']-df['low5'])
#s.reindex([1, 2, 3])
columns = df.columns.values.tolist()
#df[columns[index]]
#df = pd.DataFrame(np.random.randn(8, 4),index=dates, columns=['A', 'B', 'C', 'D'])
df = df.loc[:, ('Date','Open','High','Low', 'Close','Volume','VWAP','K','RSI', 'ZONE')]
#df = df.reindex(['Date','Open','High','Low', 'Close','Volume','VWAP','K','RSI', 'ZONE'])

df['RSI'] = calculate_rsi(df)

filter_Z1 = df['K'] <=0.1
filter_Z2 = (df['K'] > 0.1) & (df['K'] <= 0.2)
filter_Z3 = (df['K'] > 0.2) & (df['K'] <= 0.3)
filter_Z4 = (df['K'] > 0.3) & (df['K'] <= 0.4)
filter_Z5 = (df['K'] > 0.4) & (df['K'] <= 0.5)
filter_Z6 = (df['K'] > 0.5) & (df['K'] <= 0.6)
filter_Z7 = (df['K'] > 0.6) & (df['K'] <= 0.7)
filter_Z8 = (df['K'] > 0.7) & (df['K'] <= 0.8)
filter_Z9 = (df['K'] > 0.8) & (df['K'] <= 0.9)
filter_Z10 = (df['K'] > 0.9) & (df['K'] <= 1)
#plug in stochastic zones


df['ZONE'].where(-filter_Z1, 'Z1', inplace=True)
df['ZONE'].where(-filter_Z2, 'Z2', inplace=True)
df['ZONE'].where(-filter_Z3, 'Z3', inplace=True)
df['ZONE'].where(-filter_Z4, 'Z4', inplace=True)
df['ZONE'].where(-filter_Z5, 'Z5', inplace=True)
df['ZONE'].where(-filter_Z6, 'Z6', inplace=True)
df['ZONE'].where(-filter_Z7, 'Z7', inplace=True)
df['ZONE'].where(-filter_Z8, 'Z9', inplace=True)
df['ZONE'].where(-filter_Z9, 'Z9', inplace=True)
df['ZONE'].where(-filter_Z10, 'Z10', inplace=True)

df = df['Date','Open','High','Low', 'Close','Volume','VWAP','K','RSI', 'ZONE']
return df

data = yf.download('ba', period='500d', interval='1d')
df = convert_to_dataframe_daily(data)
print(df)

Answer 1

有几行需要调整

1. 而不是

df = df.loc[:, ('Date','Open','High','Low', 'Close','Volume','VWAP','K','RSI', 'ZONE')]    

使用

df = df[['Date','Open','High','Low', 'Close','Volume','VWAP','K']]

2. 之前

df['ZONE'].where(-filter_Z1, 'Z1', inplace=True)
...

放置一条线

df['ZONE'] = 0

3. return df之前的行应更改为

df = df[['Date','Open','High','Low', 'Close','Volume','VWAP','K','RSI', 'ZONE']]