QuantConnect Community

I just put the entire code:

1). I still can't see line 89.

2). I see 'd' in line 47. This worked a few minutes ago and now this seems to be a problem.

Can anyone help me on this?  Thanks :) 

The lines aren't accurate, don't read into them. I think it's because they are treating the code like C# in a python wrapper. The debugger is looking at the C# code that your python code generates, not your code itself.

One thing to keep in mind, whitespace REALLY counts on QC, a missing tab or some spaces instead of tabs can cause your code to be processed incorrectly. If your copy/paste is to be belived, you have two tabs on line 47 and 8x spaces on line 49. So the code might not process line 47 before line 49 and now you have a mysterious error that looks impossible.

Coming from Quantopian I had to copy my code into Notepad++ and enable "view whitespace characters" this allowed me to see the tabs and spaces and convert all spaces to tabs and make sure that there were no blank lines without the proper number of tabs.

example:

Hey HanByul P, LukeI  -- We just fixed the Python error line issue. If you have an example of the lines not matching again please let us know..

@Lukel, Thanks a lot ! I edited my code in the 'Notepad', copied, and pasted it in the 'main.py'. It works perfectly fine now.

@Jared, I will test this with other algo. Thanks a lot. :)

FWIW, I am having the exact same problem, on a class that is just a few lines, that I was able to run just a minute ago. I tried deleting everything and then copying it back from notepad, but to no avail. This is so frustrating....

Code follows:

import clr
clr.AddReference("System")
clr.AddReference("QuantConnect.Algorithm")
clr.AddReference("QuantConnect.Indicators")
clr.AddReference("QuantConnect.Common")

from System import *
from QuantConnect import *
from QuantConnect.Algorithm import *
from QuantConnect.Indicators import *
import numpy as np
from datetime import timedelta
from sklearn.linear_model import LinearRegression

class OptionsTest(QCAlgorithm):
    
    def Initialize(self):
        # Set the cash we'd like to use for our backtest
        # This is ignored in live trading
        self.SetCash(100000)
        
        # Start and end dates for the backtest.
        # These are ignored in live trading.
        self.SetStartDate(2014,6,30)
        self.SetEndDate(2014,7,1)
        
        # Add assets you'd like to see
        self.AddEquity("SPY", Resolution.Daily)
        self.AddEquity("XIV", Resolution.Daily)
        option = self.AddOption("SPY", Resolution.Daily)
        option.SetFilter(-2, 2, timedelta(0), timedelta(182))
        self.o_symbol=option.Symbol    

    ## we sort the contracts to find at the money (ATM) contract with farthest expiration
    def OnData(self, slice):
        d1=self.History('SPY',21,Resolution.Daily)
        d2=self.History('XIV',21,Resolution.Daily)
        index=[]
        for slice1,slice2 in zip(d1,d2):
            index.append(slice1.Time.date())
            data.append([np.float(slice1.Close),np.float(slice2.Close)])
        df=pd.DataFrame(data,columns=['SPY','XIV'],index=pd.Series(index))
        rr=df.diff().fillna(0)
        lr=LinearRegression()
        X=np.column_stack((rr['SPY'].values,rr['SPY'].values**2))
        y=rr['XIV'].values
        lr.fit(X,y)
        resid=y-lr.predict(X)
        resid=resid/resid.std()
        for kvp in slice.OptionChains:
            if kvp.Key != self.o_symbol:continue
            chain = kvp.Value
            contracts = sorted(sorted(chain,key = lambda x: abs(chain.Underlying.Price - x.Strike)),
                               key = lambda x: x.Expiry, reverse=True)
            if len(contracts) == 0: continue
            symbol = contracts[0].Symbol
            if resid[-1]>1:
                self.MarketOnCloseOrder(symbol, -1)
            if resid[-1]<-1:
                self.MarketOnCloseOrder(symbol,  1)

Copying and pasting your code Karen Chaltikian - 

The second bottom line on the stack trace is OnData -- this means the error is in OnData. Looking down the code there's no definition of "data":

Perhaps by putting it into the forum post it converted it all to spaces & made it so I could just copy-paste?

Thanks very much, Jared. This is gotta be the most creative use of forum :) It really works - I just keep a forum post window open, and any time I get any of these "Unexpected indent" or "line 0:column 0" errors, I just CtrlX into that window and back, and they're all gone.

lol! We'll look into what its doing. Maybe we just save the files in the backend without tabs (convert everything to spaces client side).

Gustavo came up with a beautiful solution for this =) We'll merge it tomorrow... 

We pushed a potential solution to this last night -- we actively convert all hard tabs to soft ones with spaces; and when we detect there's an indentation error we automatically turn on special symbols so you can see where the indentation lines up. Let us know if it solves it for you.

@Jared, Thanks. You finally fixed. At QC, now we don't have to write Python codes in 'Notepad', and then copy & paste it to the main.py. We just write our algos in the main.py as we normaly do with Python. Thanks :)