数据框将新的系列列附加到数据

Question

I have a Panda DataFrame structure and I want to add another column to it, but I can't do it with append, add or insert.

I'm trying to replicate the portfolio data with the Panda's built-in function, because this script doesn't give me correct data if the period that I request is lower than ~ 1,5 years while data must be obtained even for two days if I want. So here's the script that I want to rewrite:

import QSTK.qstkutil.qsdateutil as du
import QSTK.qstkutil.tsutil as tsu
import QSTK.qstkutil.DataAccess as da

import datetime as dt
import matplotlib.pyplot as plt
import pandas as pd

ls_symbols = ["AAPL", "GLD", "GOOG", "$SPX", "XOM"]
dt_start = dt.datetime(2006, 1, 1)
dt_end = dt.datetime(2010, 12, 31)
dt_timeofday = dt.timedelta(hours=16)
ldt_timestamps = du.getNYSEdays(dt_start, dt_end, dt_timeofday)

c_dataobj = da.DataAccess('Yahoo')
ls_keys = ['open', 'high', 'low', 'close', 'volume', 'actual_close']
ldf_data = c_dataobj.get_data(ldt_timestamps, ls_symbols, ls_keys)
**d_data = dict(zip(ls_keys, ldf_data))**

d_data = dict(zip(ls_keys, ldf_data)) is what I want to replicate because it doesn't fetch the data that I want, but I need to figure out a way to append a new column to my dict. Here is my script:

from pandas.io.data import DataReader, DataFrame
import QSTK.qstkutil.qsdateutil as du
import QSTK.qstkutil.DataAccess as da

import datetime as dt

def get_historical_data(symbol, source, date_from, date_to):
    global data_validator
    symbol_data = {}

    ls_keys = ['Open', 'High', 'Low', 'Close', 'Volume', 'Adj Close']

    for key in ls_keys:
        symbol_data[key] = DataFrame({})

    dataframe_open = DataFrame({})

    for item in symbol:
        print 'Fetching data for:', item
        current_data = DataReader(str(item), source, date_from, date_to)
        dataframe_open = {item : current_data['Open']}
        if len(symbol_data['Open'].columns) == 0:
            symbol_data['Open'] = DataFrame(dataframe_open)
        else:
            **#i want to add the new column here but can't seem to do this.**
            #symbol_data['Open'].loc[:item] = DataFrame(dataframe_open)
            pass
    return symbol_data

PS I call the func with these parameters for testing purposes:

test = get_historical_data(['SPY', 'DIA'], 'yahoo', datetime(2015,1,1), datetime(2015,1,31))

Answer 1

Does the following help? Have not tested yet, but should work in principle... Just put the data in arrays of equal length and construct the data frame from that.

def get_historical_data(symbols=[], source=None, date_from=None, date_to=None):
    global data_validator
    symbol_data = {}

    ls_keys = ['Open', 'High', 'Low', 'Close', 'Volume', 'Adj Close']
    data = []
    for item in ls_keys:
        data.append(DataReader(str(item), source, date_from, date_to)
    symbol_dataframe=DataFrame(data=data, columns=ls_keys)
    #symbol_dataframe = DataFrame()
    #for key in ls_keys:
    #    symbol_data[key] = DataFrame({})

    #dataframe_open = DataFrame({})

    #for item in symbols:
    '''    print 'Fetching data for:', item
        current_data = DataReader(str(item), source, date_from, date_to)
        dataframe_open = {item : current_data['Open']}
        #print(dataframe_open)
        if len(symbol_data['Open'].columns) == 0:
            symbol_data['Open'] = DataFrame(dataframe_open)
        else:
            #i want to add the new column here but can't seem to do this.**
            symbol_data['Open'] = DataFrame(dataframe_open)
            symbol_data.head()
    '''
    return symbol_dataframe