from quantopian.pipeline import Pipeline
from quantopian.pipeline.data import USEquityPricing
from quantopian.pipeline.experimental import QTradableStocksUS
from quantopian.research import run_pipeline
from quantopian.pipeline.data import Fundamentals

def make_pipeline():
    base_universe = QTradableStocksUS()
    close_price = USEquityPricing.close.latest
    
    #Profitability Ratios
    earning_yield = Fundamentals.earning_yield.latest
    return_on_equity = Fundamentals.roe.latest
    return_on_assets = Fundamentals.roa.latest
    net_profit = Fundamentals.net_margin.latest
    
    #Liquidity Ratios
    quick_ratio = Fundamentals.quick_ratio.latest
    current_ratio = Fundamentals.current_ratio.latest
    debt_to_equity_ratio = (Fundamentals.long_term_debt_equity_ratio.latest)**-1
    interest_coverage = Fundamentals.interest_coverage.latest
    
    #Efficiency Ratios
    assets_turnover = Fundamentals.assets_turnover.latest
    change_in_working_capital = Fundamentals.change_in_working_capital.latest
    
    #Investment Ratios
    book_value_yield = Fundamentals.book_value_yield.latest
    fcf_yield = Fundamentals.fcf_yield.latest
    trailing_dividend_yield = Fundamentals.trailing_dividend_yield.latest
    total_yield = Fundamentals.total_yield.latest
    
    #Winsorising
    earning_yield_winsorized = earning_yield.winsorize(min_percentile=0.01, max_percentile=0.99)
    return_on_equity_winsorized = return_on_equity.winsorize(min_percentile=0.01, max_percentile=0.99)
    return_on_assets_winsorized = return_on_assets.winsorize(min_percentile=0.01, max_percentile=0.99)
    net_profit_winsorized = net_profit.winsorize(min_percentile=0.01, max_percentile=0.99)
    quick_ratio_winsorized = quick_ratio.winsorize(min_percentile=0.01, max_percentile=0.99)
    current_ratio_winsorized = current_ratio.winsorize(min_percentile=0.01, max_percentile=0.99)
    debt_to_equity_ratio_winsorized = debt_to_equity_ratio.winsorize(min_percentile=0.01, max_percentile=0.99)
    interest_coverage_winsorized = interest_coverage.winsorize(min_percentile=0.01, max_percentile=0.99)
    assets_turnover_winsorized = assets_turnover.winsorize(min_percentile=0.01, max_percentile=0.99)
    change_in_working_capital_winsorized = change_in_working_capital.winsorize(min_percentile=0.01, max_percentile=0.99)
    book_value_yield_winsorized = book_value_yield.winsorize(min_percentile=0.01, max_percentile=0.99)
    fcf_yield_winsorized = fcf_yield.winsorize(min_percentile=0.01, max_percentile=0.99)
    trailing_dividend_yield_winsorized = trailing_dividend_yield.winsorize(min_percentile=0.01, max_percentile=0.99)
    total_yield_winsorized = total_yield.winsorize(min_percentile=0.01, max_percentile=0.99)
    
    #Creating a combined factor


    return Pipeline(
    columns = {
        'close price': close_price,
        'earning_yield': earning_yield_winsorized.zscore(),
        'return_on_equity': return_on_equity_winsorized.zscore(),
        'return_on_assets': return_on_assets_winsorized.zscore(),
        'net_profit': net_profit_winsorized.zscore(),
        'quick_ratio': quick_ratio_winsorized.zscore(),
        'current_ratio': current_ratio_winsorized.zscore(),
        'debt_to_equity_ratio': debt_to_equity_ratio_winsorized.zscore(),
        'interest_coverage': interest_coverage_winsorized.zscore(),
        'assets_turnover': assets_turnover_winsorized.zscore(),
        'change_in_working_capital': change_in_working_capital_winsorized.zscore(),
        'book_value_yield': book_value_yield_winsorized.zscore(),
        'fcf_yield': fcf_yield_winsorized.zscore(),
        'trailing_dividend_yield': trailing_dividend_yield_winsorized.zscore(),
        'total_yield': total_yield_winsorized.zscore(),
    },
    screen = (base_universe),)

period_start = '2009-01-01'
period_end = '2014-01-01'

pipeline_output = run_pipeline(
    make_pipeline(),
    start_date = period_start,
    end_date = period_end)

pipeline_output.info()

<class 'pandas.core.frame.DataFrame'>
MultiIndex: 2299567 entries, (2009-01-02 00:00:00+00:00, Equity(2 [ARNC])) to (2014-01-02 00:00:00+00:00, Equity(44510 [FWM]))
Data columns (total 15 columns):
assets_turnover              float64
book_value_yield             float64
change_in_working_capital    float64
close price                  float64
current_ratio                float64
debt_to_equity_ratio         float64
earning_yield                float64
fcf_yield                    float64
interest_coverage            float64
net_profit                   float64
quick_ratio                  float64
return_on_assets             float64
return_on_equity             float64
total_yield                  float64
trailing_dividend_yield      float64
dtypes: float64(15)
memory usage: 280.7+ MB

pipeline_output.dropna(inplace=True)

pipeline_output.info()

<class 'pandas.core.frame.DataFrame'>
MultiIndex: 939317 entries, (2009-01-02 00:00:00+00:00, Equity(41 [ARCB])) to (2014-01-02 00:00:00+00:00, Equity(44508 [CST]))
Data columns (total 15 columns):
assets_turnover              939317 non-null float64
book_value_yield             939317 non-null float64
change_in_working_capital    939317 non-null float64
close price                  939317 non-null float64
current_ratio                939317 non-null float64
debt_to_equity_ratio         939317 non-null float64
earning_yield                939317 non-null float64
fcf_yield                    939317 non-null float64
interest_coverage            939317 non-null float64
net_profit                   939317 non-null float64
quick_ratio                  939317 non-null float64
return_on_assets             939317 non-null float64
return_on_equity             939317 non-null float64
total_yield                  939317 non-null float64
trailing_dividend_yield      939317 non-null float64
dtypes: float64(15)
memory usage: 114.7+ MB

pipeline_output['30d_future_close'] = pipeline_output['close price'].shift(-30)
pipeline_output['30d_close_future_pct'] = pipeline_output['30d_future_close'].pct_change(30)

pipeline_output.dropna(inplace=True)

pipeline_output.head()

pipeline_output.info()

<class 'pandas.core.frame.DataFrame'>
MultiIndex: 939257 entries, (2009-01-02 00:00:00+00:00, Equity(3490 [HCP])) to (2014-01-02 00:00:00+00:00, Equity(41242 [ARCO]))
Data columns (total 17 columns):
assets_turnover              939257 non-null float64
book_value_yield             939257 non-null float64
change_in_working_capital    939257 non-null float64
close price                  939257 non-null float64
current_ratio                939257 non-null float64
debt_to_equity_ratio         939257 non-null float64
earning_yield                939257 non-null float64
fcf_yield                    939257 non-null float64
interest_coverage            939257 non-null float64
net_profit                   939257 non-null float64
quick_ratio                  939257 non-null float64
return_on_assets             939257 non-null float64
return_on_equity             939257 non-null float64
total_yield                  939257 non-null float64
trailing_dividend_yield      939257 non-null float64
30d_future_close             939257 non-null float64
30d_close_future_pct         939257 non-null float64
dtypes: float64(17)
memory usage: 129.0+ MB

# Import Lasso
from sklearn.linear_model import Lasso
import matplotlib.pyplot as plt

# Instantiate a lasso regressor: lasso
lasso = Lasso(alpha=0.4,normalize=True)

X = pipeline_output['30d_close_future_pct']
y = pipeline_output[pipeline_output.columns[~pipeline_output.columns.isin(['30d_future_close','30d_close_future_pct','close price'])]]
y = y.values
X = X.values

lasso.fit(X,y)

lasso_coef = lasso.coef_


ValueErrorTraceback (most recent call last)
<ipython-input-77-75f11c6f87ae> in <module>()
----> 1 lasso.fit(X,y)
      2 
      3 lasso_coef = lasso.coef_

/usr/local/lib/python2.7/dist-packages/sklearn/linear_model/coordinate_descent.pyc in fit(self, X, y)
    628         X, y = check_X_y(X, y, accept_sparse='csc', dtype=np.float64,
    629                          order='F', copy=self.copy_X and self.fit_intercept,
--> 630                          multi_output=True, y_numeric=True)
    631 
    632         X, y, X_mean, y_mean, X_std, precompute, Xy = \

/usr/local/lib/python2.7/dist-packages/sklearn/utils/validation.pyc in check_X_y(X, y, accept_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric)
    452         y = y.astype(np.float64)
    453 
--> 454     check_consistent_length(X, y)
    455 
    456     return X, y

/usr/local/lib/python2.7/dist-packages/sklearn/utils/validation.pyc in check_consistent_length(*arrays)
    172     if len(uniques) > 1:
    173         raise ValueError("Found arrays with inconsistent numbers of samples: "
--> 174                          "%s" % str(uniques))
    175 
    176 

ValueError: Found arrays with inconsistent numbers of samples: [     1 939257]

		assets_turnover	book_value_yield	change_in_working_capital	close price	current_ratio	debt_to_equity_ratio	earning_yield	fcf_yield	interest_coverage	net_profit	quick_ratio	return_on_assets	return_on_equity	total_yield	trailing_dividend_yield	30d_future_close	30d_close_future_pct
2009-01-02 00:00:00+00:00	Equity(3490 [HCP])	-0.906487	-0.299722	-0.073550	27.78	0.255053	-0.181476	0.290604	0.279296	-0.127925	0.217826	0.462359	0.285918	0.223411	-0.122917	-0.150549	26.59	-0.042837
	Equity(3519 [HEI])	0.187917	-0.471460	-0.105106	38.82	0.029198	-0.176972	0.298925	0.282467	0.264632	0.176252	-0.325887	0.433042	0.238756	-0.639812	-0.462101	47.86	0.232870
	Equity(3596 [HMA])	0.172728	-0.469821	-0.365603	1.82	-0.240964	-0.181870	0.506210	0.650853	-0.132968	0.166823	-0.172285	0.165487	0.226685	-0.661178	-0.431391	53.60	28.450549
	Equity(3668 [HRL])	1.666380	-0.428789	-0.157392	31.03	-0.232483	-0.178946	0.308525	0.259000	-0.052672	0.170366	-0.443837	0.412780	0.240023	-0.329182	-0.357094	18.06	-0.417983
	Equity(3676 [HRS])	0.511444	-0.465372	-0.349073	38.04	-0.285473	-0.180500	0.305599	0.306674	-0.036738	0.176215	-0.315335	0.508437	0.276552	-0.184550	-0.383841	24.66	-0.351735