Pandas 在从 csv 文件读取列时出错

Question

So I have a google stock dataset that has a 'Date', 'Open', 'High', 'Low', 'Close' and 'Volume' columns. The problem is that pandas keeps saying that the 'Close' column is of type 'Object'.

I found the problem when I tried to scale that column using the MinMaxScaler and that's when I got this error: could not convert string to float: '1,008.64'

When I tried converting it to a float using the astype('float') function, I get the same error. When I try the to_numeric(price['Close'], errors='coerce') function, it just sets some rows as NULL.

My code:

import pandas as pd

train_data = pd.read_csv('drive/MyDrive/Data LSTM TimeSeries/Google_Stock_Price.csv')

price = train_data[['Close']]
price.info()

from sklearn.preprocessing import MinMaxScaler

scaler = MinMaxScaler(feature_range=(-1,1))
price['Close'] = scaler.fit_transform(price['Close'].values.reshape(-1,1))

price['Close'] = price['Close'].astype('float')
price['Close'] = pd.to_numeric(price['Close'], errors='ignore')

Answer 1

Try using the thousands argument when you use read_csv to make sure that your values are read correctly with the , delimiter

train_data = pd.read_csv('drive/MyDrive/Data LSTM TimeSeries/Google_Stock_Price.csv', thousands = ",")