def test_get_top_n(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'prev_returns': pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051]], dates, tickers), 'top_n': 3} fn_correct_outputs = OrderedDict([ ( 'top_stocks', pd.DataFrame( [ [0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [1, 0, 1, 1, 0], [0, 1, 0, 1, 1]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_positions(fn): tickers = generate_random_tickers(5) dates = generate_random_dates(6) fn_inputs = { 'prices': pd.DataFrame( [ [65.40757705426432, 27.556319958924323, 50.59935209411175, 56.274712269629134, 99.9873070881051], [47.82126720752715, 56.812865745668375, 40.75685814634723, 27.469680989736023, 41.449858088448735], [88.20038097315815, 45.573972499280494, 36.592711369868724, 21.36570423559795, 0.698919959739297], [14.670236824202721, 49.557949251949054, 18.935364730808935, 23.163368660093298, 8.075599541367884], [41.499140208637705, 9.75987296846733, 66.08677766062186, 37.927861417544385, 10.792730405945827], [86.26923464863536, 32.12679487375028, 15.621592524570282, 77.1908860965619, 52.733950486350444]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'final_positions', pd.DataFrame( [ [30, 0, 30, 30, 30], [0, 30, 0, 0, 0], [30, 0, 0, 0, -10], [-10, 0, -10, 0, -10], [0, -10, 30, 0, -10], [30, 0, -10, 30, 30]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_top_n(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'prev_returns': pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051]], dates, tickers), 'top_n': 3} fn_correct_outputs = OrderedDict([ ( 'top_stocks', pd.DataFrame( [ [0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [1, 0, 1, 1, 0], [0, 1, 0, 1, 1]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_resample_prices(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-19', '2008-09-08', '2008-09-28', '2008-10-18', '2008-11-07', '2008-11-27']) resampled_dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'close_prices': pd.DataFrame( [ [21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235], [15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022], [482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077], [10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 10.52435923388642], [10.675971965144655, 12.749401436636365, 11.805257579935713, 21.539039489843024, 19.99766036804861], [11.545495378369814, 23.981468434099405, 24.974763062186504, 36.031962102997689, 14.304332320024963]], dates, tickers), 'freq': 'M'} fn_correct_outputs = OrderedDict([ ( 'prices_resampled', pd.DataFrame( [ [21.05081048, 17.01384381, 10.98450376, 11.24809343, 12.96171273], [482.34539247, 35.20258059, 3516.54167823, 66.40531433, 13.50396048], [10.91893302, 17.90864387, 24.80126542, 12.48895419, 10.52435923], [11.54549538, 23.98146843, 24.97476306, 36.03196210, 14.30433232]], resampled_dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_resample_prices(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-19', '2008-09-08', '2008-09-28', '2008-10-18', '2008-11-07', '2008-11-27']) resampled_dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'close_prices': pd.DataFrame( [ [21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235], [15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022], [482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077], [10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 10.52435923388642], [10.675971965144655, 12.749401436636365, 11.805257579935713, 21.539039489843024, 19.99766036804861], [11.545495378369814, 23.981468434099405, 24.974763062186504, 36.031962102997689, 14.304332320024963]], dates, tickers), 'freq': 'M'} fn_correct_outputs = OrderedDict([ ( 'prices_resampled', pd.DataFrame( [ [21.05081048, 17.01384381, 10.98450376, 11.24809343, 12.96171273], [482.34539247, 35.20258059, 3516.54167823, 66.40531433, 13.50396048], [10.91893302, 17.90864387, 24.80126542, 12.48895419, 10.52435923], [11.54549538, 23.98146843, 24.97476306, 36.03196210, 14.30433232]], resampled_dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_shift_returns(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'returns': pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995]], dates, tickers), 'shift_n': 1} fn_correct_outputs = OrderedDict([ ( 'shifted_returns', pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_high_lows_lookback(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'high': pd.DataFrame([[35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107]], dates, tickers), 'low': pd.DataFrame([[15.6718, 75.1392, 34.0527], [27.1834, 12.3453, 95.9373], [28.2503, 24.2854, 23.2932], [86.3725, 32.223, 38.4107]], dates, tickers), 'lookback_days': 2 } fn_correct_outputs = OrderedDict([ ('lookback_high', pd.DataFrame([[np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan], [92.11310000, 91.05430000, 90.95720000], [92.11310000, 91.05430000, 90.95720000]], dates, tickers)), ('lookback_low', pd.DataFrame([[np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan], [15.67180000, 12.34530000, 34.05270000], [27.18340000, 12.34530000, 23.29320000]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_date_top_industries(fn=date_top_industries): tickers = generate_random_tickers(10) dates = generate_random_dates(2) fn_inputs = { 'prices': pd.DataFrame( [[ 21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235, 482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077 ], [ 15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022, 10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 15.63570258751384 ]], dates, tickers), 'sector': pd.Series([ 'ENERGY', 'MATERIALS', 'ENERGY', 'ENERGY', 'TELECOM', 'FINANCIALS', 'TECHNOLOGY', 'HEALTH', 'MATERIALS', 'REAL ESTATE' ], tickers), 'date': dates[-1], 'top_n': 4 } fn_correct_outputs = OrderedDict([('top_industries', {'ENERGY', 'HEALTH', 'TECHNOLOGY'})]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { "prices": pd.DataFrame( [ [35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107], ], dates, tickers, ) } fn_correct_outputs = OrderedDict( [ ( "returns", pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958], ], dates, tickers, ), ) ] ) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_dollar_volume_weights(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'close': pd.DataFrame( [ [35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107]], dates, tickers), 'volume': pd.DataFrame( [ [9.83683e+06, 1.78072e+07, 8.82982e+06], [8.22427e+07, 6.85315e+07, 4.81601e+07], [1.62348e+07, 1.30527e+07, 9.51201e+06], [1.06742e+07, 5.68313e+07, 9.31601e+06]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'dollar_volume_weights', pd.DataFrame( [ [0.27719777, 0.48394253, 0.23885970], [0.41632975, 0.34293308, 0.24073717], [0.41848548, 0.33536102, 0.24615350], [0.05917255, 0.85239760, 0.08842984]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_long_short(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'close': pd.DataFrame([[25.6788, 35.1392, 34.0527], [25.1884, 14.3453, 39.9373], [78.2803, 34.3854, 23.2932], [88.8725, 52.223, 34.4107]], dates, tickers), 'lookback_high': pd.DataFrame( [[np.nan, np.nan, np.nan], [92.11310000, 91.05430000, 90.95720000], [35.4411, 34.1799, 34.0223], [92.11310000, 91.05430000, 90.95720000]], dates, tickers), 'lookback_low': pd.DataFrame([[np.nan, np.nan, np.nan], [34.1705, 92.453, 58.5107], [15.67180000, 12.34530000, 34.05270000], [27.18340000, 12.34530000, 23.29320000]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('long_short', pd.DataFrame([[0, 0, 0], [-1, -1, -1], [1, 1, -1], [0, 0, 0]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_weighted_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'returns': pd.DataFrame( [[np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958]], dates, tickers), 'weights': pd.DataFrame([[0.03777059, 0.04733924, 0.05197790], [0.82074874, 0.48533938, 0.75792752], [0.10196420, 0.05866016, 0.09578226], [0.03951647, 0.40866122, 0.09431233]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('weighted_returns', pd.DataFrame( [[np.nan, np.nan, np.nan], [1.31241616, 0.80759119, 1.26836009], [-0.03779367, -0.02143549, -0.03449679], [-0.01622375, 0.24521625, 0.00050642]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_portfolio_returns(fn=utils.portfolio_returns): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex( ['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'df_long': pd.DataFrame([[0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [1, 0, 1, 1, 0], [0, 1, 0, 1, 1]], dates, tickers), 'df_short': pd.DataFrame([[0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [0, 1, 0, 1, 1], [1, 1, 1, 0, 0]], dates, tickers), 'lookahead_returns': pd.DataFrame( [[3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995], [1.25459098, 6.87369275, 2.58265839, 6.92676837, 0.84632677]], dates, tickers), 'n_stocks': 3 } fn_correct_outputs = OrderedDict([ ('portfolio_returns', pd.DataFrame([ [0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00000000], [-0.00000000, -0.00000000, -0.00000000, -0.00000000, -0.00000000], [0.01859903, -0.09733263, 0.00232372, 0.00000000, -0.10228998], [-0.41819699, 0.00000000, -0.86088613, 2.30892279, 0.28210892] ], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_signal_return(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(5) fn_inputs = { 'signal': pd.DataFrame( [[0, 0, 0], [-1, -1, -1], [1, 0, 0], [0, 0, 0], [0, 1, 0]], dates, tickers), 'lookahead_returns': pd.DataFrame( [[0.88702896, 0.96521098, 0.65854789], [1.13391240, 0.87420969, -0.53914925], [0.35450805, -0.56900529, -0.64808965], [0.38572896, -0.94655617, 0.123564379], [np.nan, np.nan, np.nan]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('signal_return', pd.DataFrame( [[0, 0, 0], [-1.13391240, -0.87420969, 0.53914925], [0.35450805, 0, 0], [0, 0, 0], [np.nan, np.nan, np.nan]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_lookahead_prices(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(5) fn_inputs = { 'close': pd.DataFrame( [ [25.6788, 35.1392, 34.0527], [25.1884, 14.3453, 39.9373], [62.3457, 92.2524, 65.7893], [78.2803, 34.3854, 23.2932], [88.8725, 52.223, 34.4107]], dates, tickers), 'lookahead_days': 2} fn_correct_outputs = OrderedDict([ ( 'lookahead_prices', pd.DataFrame( [ [62.34570000, 92.25240000, 65.78930000], [78.28030000, 34.38540000, 23.29320000], [88.87250000, 52.22300000, 34.41070000], [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_return_lookahead(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(5) fn_inputs = { 'close': pd.DataFrame( [ [25.6788, 35.1392, 34.0527], [25.1884, 14.3453, 39.9373], [62.3457, 92.2524, 65.7893], [78.2803, 34.3854, 23.2932], [88.8725, 52.223, 34.4107]], dates, tickers), 'lookahead_prices': pd.DataFrame( [ [62.34570000, 92.25240000, 65.78930000], [78.28030000, 34.38540000, 23.29320000], [88.87250000, 52.22300000, 34.41070000], [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'lookahead_returns', pd.DataFrame( [ [0.88702896, 0.96521098, 0.65854789], [1.13391240, 0.87420969, -0.53914925], [0.35450805, -0.56900529, -0.64808965], [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_dollar_volume_weights(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'close': pd.DataFrame( [ [35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107]], dates, tickers), 'volume': pd.DataFrame( [ [9.83683e+06, 1.78072e+07, 8.82982e+06], [8.22427e+07, 6.85315e+07, 4.81601e+07], [1.62348e+07, 1.30527e+07, 9.51201e+06], [1.06742e+07, 5.68313e+07, 9.31601e+06]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'dollar_volume_weights', pd.DataFrame( [ [0.27719777, 0.48394253, 0.23885970], [0.41632975, 0.34293308, 0.24073717], [0.41848548, 0.33536102, 0.24615350], [0.05917255, 0.85239760, 0.08842984]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_weighted_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'returns': pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958]], dates, tickers), 'weights': pd.DataFrame( [ [0.03777059, 0.04733924, 0.05197790], [0.82074874, 0.48533938, 0.75792752], [0.10196420, 0.05866016, 0.09578226], [0.03951647, 0.40866122, 0.09431233]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'weighted_returns', pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.31241616, 0.80759119, 1.26836009], [-0.03779367, -0.02143549, -0.03449679], [-0.01622375, 0.24521625, 0.00050642]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_returns(fn): tickers = generate_random_tickers(5) dates = generate_random_dates(6) fn_inputs = { 'close': pd.DataFrame( [ [21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235], [15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022], [482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077], [10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 10.52435923388642], [10.675971965144655, 12.749401436636365, 11.805257579935713, 21.539039489843024, 19.99766036804861], [11.545495378369814, 23.981468434099405, 24.974763062186504, 36.031962102997689, 14.304332320024963]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'returns', pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [-0.25723988, -0.13655355, 0.03354944, 41.29534136, -0.07731018], [29.84897463, 1.39627496, 308.74483411, -0.86041737, 0.12912763], [-0.97736283, -0.49126900, -0.99294726, -0.81192839, -0.22064647], [-0.02225135, -0.28808672, -0.52400584, 0.72464717, 0.90013092], [0.08144677, 0.88098779, 1.11556274, 0.67286764, -0.28469971]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_kstest(fn): tickers = generate_random_tickers(3) fn_inputs = { "long_short_signal_returns": pd.DataFrame({ "ticker": tickers * 5, "signal_return": [ 0.12, -0.83, 0.37, 0.83, -0.34, 0.27, -0.68, 0.29, 0.69, 0.57, 0.39, 0.56, -0.97, -0.72, 0.26, ], }) } fn_correct_outputs = OrderedDict([ ("ks_values", pd.Series([0.29787827, 0.35221525, 0.63919407], tickers)), ("p_values", pd.Series([0.67234149, 0.46161172, 0.01650327], tickers)), ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_simple_moving_average(fn): tickers = generate_random_tickers(5) dates = generate_random_dates(6) fn_inputs = { 'close': pd.DataFrame( [ [21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235], [15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022], [482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077], [10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 10.52435923388642], [10.675971965144655, 12.749401436636365, 11.805257579935713, 21.539039489843024, 19.99766036804861], [11.545495378369814, 23.981468434099405, 24.974763062186504, 36.031962102997689, 14.304332320024963]], dates, tickers), 'rolling_window': 3} fn_correct_outputs = OrderedDict([ ( 'returns', pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan, np.nan, np.nan], [173.01063518,22.30232250,1179.62640322,184.46511965,12.80843788], [169.63334269,22.60058918,1184.23199044,184.87873990,11.99598671], [167.98009915,21.95354197,1184.38273374,33.47776934,14.67532669], [11.04680012,18.21317125,20.52709535,23.35331859,14.94211731]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_shift_returns(fn=utils.shift_returns): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex( ['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'returns': pd.DataFrame( [[np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995]], dates, tickers), 'shift_n': 1 } fn_correct_outputs = OrderedDict([ ('shifted_returns', pd.DataFrame([ [np.nan, np.nan, np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051] ], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_return_lookahead(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(5) fn_inputs = { 'close': pd.DataFrame([[25.6788, 35.1392, 34.0527], [25.1884, 14.3453, 39.9373], [62.3457, 92.2524, 65.7893], [78.2803, 34.3854, 23.2932], [88.8725, 52.223, 34.4107]], dates, tickers), 'lookahead_prices': pd.DataFrame([[62.34570000, 92.25240000, 65.78930000], [78.28030000, 34.38540000, 23.29320000], [88.87250000, 52.22300000, 34.41070000], [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('lookahead_returns', pd.DataFrame([[0.88702896, 0.96521098, 0.65854789], [1.13391240, 0.87420969, -0.53914925], [0.35450805, -0.56900529, -0.64808965], [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_compute_log_returns(fn=utils.compute_log_returns): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex( ['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'prices': pd.DataFrame([[ 21.05081048, 17.01384381, 10.98450376, 11.24809343, 12.96171273 ], [ 482.34539247, 35.20258059, 3516.54167823, 66.40531433, 13.50396048 ], [ 10.91893302, 17.90864387, 24.80126542, 12.48895419, 10.52435923 ], [11.54549538, 23.98146843, 24.97476306, 36.03196210, 14.30433232]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('log_returns', pd.DataFrame([ [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995] ], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_covariance_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { "returns": pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958], ], dates, tickers, ) } fn_correct_outputs = OrderedDict( [ ( "returns_covariance", np.array( [ [0.89856076, 0.7205586, 0.8458721], [0.7205586, 0.78707297, 0.76450378], [0.8458721, 0.76450378, 0.83182775], ] ), ) ] ) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_dividend_weights(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { "dividends": pd.DataFrame( [[0.0, 0.0, 0.0], [0.0, 0.0, 0.1], [0.0, 1.0, 0.3], [0.0, 0.2, 0.0]], dates, tickers, ) } fn_correct_outputs = OrderedDict( [ ( "dividend_weights", pd.DataFrame( [ [np.nan, np.nan, np.nan], [0.00000000, 0.00000000, 1.00000000], [0.00000000, 0.71428571, 0.28571429], [0.00000000, 0.75000000, 0.25000000], ], dates, tickers, ), ) ] ) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_cumulative_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { "returns": pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958], ], dates, tickers, ) } fn_correct_outputs = OrderedDict( [ ( "cumulative_returns", pd.Series([np.nan, 5.93647782, -0.57128454, -0.68260542], dates), ) ] ) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_signal_return(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(5) fn_inputs = { 'signal': pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 0, 0], [0, 0, 0], [0, 1, 0]], dates, tickers), 'lookahead_returns': pd.DataFrame( [ [0.88702896, 0.96521098, 0.65854789], [1.13391240, 0.87420969, -0.53914925], [0.35450805, -0.56900529, -0.64808965], [0.38572896, -0.94655617, 0.123564379], [np.nan, np.nan, np.nan]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'signal_return', pd.DataFrame( [ [0, 0, 0], [-1.13391240, -0.87420969, 0.53914925], [0.35450805, 0, 0], [0, 0, 0], [np.nan, np.nan, np.nan]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_find_outliers(fn): tickers = generate_random_tickers(3) fn_inputs = { "ks_values": pd.Series([0.20326939, 0.34826827, 0.60256811], tickers), "p_values": pd.Series([0.98593727, 0.48009144, 0.02898631], tickers), "ks_threshold": 0.5, "pvalue_threshold": 0.05, } fn_correct_outputs = OrderedDict([("outliers", set([tickers[2]]))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_find_outliers(fn=utils.find_outliers): tickers = generate_random_tickers(3) fn_inputs = { 'ks_values': pd.Series([0.20326939, 0.34826827, 0.60256811], tickers), 'p_values': pd.Series([0.98593727, 0.48009144, 0.02898631], tickers), 'ks_threshold': 0.5, 'pvalue_threshold': 0.05 } fn_correct_outputs = OrderedDict([('outliers', set([tickers[2]]))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_rebalance_portfolio(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(11) fn_inputs = { 'returns': pd.DataFrame( [[np.nan, np.nan, np.nan], [-0.02202381, 0.02265285, 0.01441961], [0.01947657, 0.00551985, 0.00047382], [0.00537313, -0.00803232, 0.01160313], [0.00593824, -0.00567773, 0.02247191], [0.02479339, 0.01758824, -0.00824176], [-0.0109447, -0.00383568, 0.01361958], [0.01164822, 0.01558719, 0.00614894], [0.0109384, -0.00182079, 0.02900868], [0.01138952, 0.00218049, -0.00954495], [0.0106982, 0.00644535, -0.01815329]], dates, tickers), 'index_weights': pd.DataFrame([[0.00449404, 0.11586048, 0.00359727], [ 0.00403487, 0.12534048, 0.0034428, ], [0.00423485, 0.12854258, 0.00347404], [0.00395679, 0.1243466, 0.00335064], [0.00368729, 0.11750295, 0.00333929], [0.00369562, 0.11447422, 0.00325973], [ 0.00379612, 0.11088075, 0.0031734, ], [0.00366501, 0.10806014, 0.00314648], [0.00361268, 0.10376514, 0.00323257], [0.00358844, 0.10097531, 0.00319009], [0.00362045, 0.09791232, 0.00318071]], dates, tickers), 'shift_size': 2, 'chunk_size': 4 } fn_correct_outputs = OrderedDict([('all_rebalance_weights', [ np.array([0.29341237, 0.41378419, 0.29280344]), np.array([0.29654088, 0.40731481, 0.29614432]), np.array([0.29868214, 0.40308791, 0.29822995]), np.array([0.30100044, 0.39839644, 0.30060312]) ])]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_rebalance_portfolio(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(11) fn_inputs = { 'returns': pd.DataFrame( [ [np.nan, np.nan, np.nan], [-0.02202381, 0.02265285, 0.01441961], [0.01947657, 0.00551985, 0.00047382], [0.00537313, -0.00803232, 0.01160313], [0.00593824, -0.00567773, 0.02247191], [0.02479339, 0.01758824, -0.00824176], [-0.0109447, -0.00383568, 0.01361958], [0.01164822, 0.01558719, 0.00614894], [0.0109384, -0.00182079, 0.02900868], [0.01138952, 0.00218049, -0.00954495], [0.0106982, 0.00644535, -0.01815329]], dates, tickers), 'index_weights': pd.DataFrame( [ [0.00449404, 0.11586048, 0.00359727], [0.00403487, 0.12534048, 0.0034428, ], [0.00423485, 0.12854258, 0.00347404], [0.00395679, 0.1243466, 0.00335064], [0.00368729, 0.11750295, 0.00333929], [0.00369562, 0.11447422, 0.00325973], [0.00379612, 0.11088075, 0.0031734, ], [0.00366501, 0.10806014, 0.00314648], [0.00361268, 0.10376514, 0.00323257], [0.00358844, 0.10097531, 0.00319009], [0.00362045, 0.09791232, 0.00318071]], dates, tickers), 'shift_size': 2, 'chunk_size': 4} fn_correct_outputs = OrderedDict([ ( 'all_rebalance_weights', [ np.array([0.29341237, 0.41378419, 0.29280344]), np.array([0.29654088, 0.40731481, 0.29614432]), np.array([0.29868214, 0.40308791, 0.29822995]), np.array([0.30100044, 0.39839644, 0.30060312])] )]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_filter_signals(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(10) fn_inputs = { "signal": pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 0, -1], [0, 0, 0], [1, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 1], [-1, 0, 0], [0, 0, 0], ], dates, tickers, ), "lookahead_days": 3, } fn_correct_outputs = OrderedDict([( "filtered_signal", pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 0, 0], [0, 0, 0], [0, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 0], [-1, 0, 0], [0, 0, 0], ], dates, tickers, ), )]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_find_outliers(fn): tickers = generate_random_tickers(3) fn_inputs = { 'ks_values': pd.Series( [0.20326939, 0.34826827, 0.60256811], tickers), 'p_values': pd.Series( [0.98593727, 0.48009144, 0.02898631], tickers), 'ks_threshold': 0.5, 'pvalue_threshold': 0.05} fn_correct_outputs = OrderedDict([ ( 'outliers', set([tickers[2]]))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_high_lows_lookback(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'high': pd.DataFrame( [ [35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107]], dates, tickers), 'low': pd.DataFrame( [ [15.6718, 75.1392, 34.0527], [27.1834, 12.3453, 95.9373], [28.2503, 24.2854, 23.2932], [86.3725, 32.223, 38.4107]], dates, tickers), 'lookback_days': 2} fn_correct_outputs = OrderedDict([ ( 'lookback_high', pd.DataFrame( [ [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan], [92.11310000, 91.05430000, 90.95720000], [92.11310000, 91.05430000, 90.95720000]], dates, tickers)), ( 'lookback_low', pd.DataFrame( [ [np.nan, np.nan, np.nan], [np.nan, np.nan, np.nan], [15.67180000, 12.34530000, 34.05270000], [27.18340000, 12.34530000, 23.29320000]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_cumulative_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'returns': pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'cumulative_returns', pd.Series( [np.nan, 5.93647782, -0.57128454, -0.68260542], dates))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_kstest(fn): tickers = generate_random_tickers(3) fn_inputs = { 'long_short_signal_returns': pd.DataFrame({ 'ticker': tickers * 5, 'signal_return': [ 0.12, -0.83, 0.37, 0.83, -0.34, 0.27, -0.68, 0.29, 0.69, 0.57, 0.39, 0.56, -0.97, -0.72, 0.26 ] }) } fn_correct_outputs = OrderedDict([ ('ks_values', pd.Series([0.28999582, 0.34484969, 0.63466098], tickers)), ('p_values', pd.Series([0.73186935, 0.49345487, 0.01775987], tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_kstest(fn=utils.calculate_kstest): tickers = generate_random_tickers(3) fn_inputs = { 'long_short_signal_returns': pd.DataFrame({ 'ticker': tickers * 5, 'signal_return': [ 0.12, -0.83, 0.37, 0.83, -0.34, 0.27, -0.68, 0.29, 0.69, 0.57, 0.39, 0.56, -0.97, -0.72, 0.26 ] }) } fn_correct_outputs = OrderedDict([ ('ks_values', pd.Series([0.29787827, 0.35221525, 0.63919407], tickers)), ('p_values', pd.Series([0.69536353, 0.46493498, 0.01650327], tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_covariance_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'returns': pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958]], dates, tickers)} fn_correct_outputs = OrderedDict([( 'returns_covariance', np.array( [ [0.89856076, 0.7205586, 0.8458721], [0.7205586, 0.78707297, 0.76450378], [0.8458721, 0.76450378, 0.83182775]]))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_arithmetic_rate_of_return(fn): tickers = generate_random_tickers(5) dates = generate_random_dates(6) fn_inputs = { 'close': pd.DataFrame( [[ 21.050810483942833, 17.013843810658827, 10.984503755486879, 11.248093428369392, 12.961712733997235 ], [ 15.63570258751384, 14.69054309070934, 11.353027688995159, 475.74195118202061, 11.959640427803022 ], [ 482.34539247360806, 35.202580592515041, 3516.5416782257166, 66.405314327318209, 13.503960481087077 ], [ 10.918933017418304, 17.9086438675435, 24.801265417692324, 12.488954191854916, 10.52435923388642 ], [ 10.675971965144655, 12.749401436636365, 11.805257579935713, 21.539039489843024, 19.99766036804861 ], [ 11.545495378369814, 23.981468434099405, 24.974763062186504, 36.031962102997689, 14.304332320024963 ]], dates, tickers) } fn_correct_outputs = OrderedDict([ ('arithmetic_returns', pd.Series( [4.77892789, 0.22689225, 51.39616553, 6.83675173, 0.07443370], tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_portfolio_returns(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'df_long': pd.DataFrame( [ [0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [1, 0, 1, 1, 0], [0, 1, 0, 1, 1]], dates, tickers), 'df_short': pd.DataFrame( [ [0, 0, 0, 0, 0], [0, 0, 0, 0, 0], [0, 1, 0, 1, 1], [1, 1, 1, 0, 0]], dates, tickers), 'lookahead_returns': pd.DataFrame( [ [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995], [1.25459098, 6.87369275, 2.58265839, 6.92676837, 0.84632677]], dates, tickers), 'n_stocks': 3} fn_correct_outputs = OrderedDict([ ( 'portfolio_returns', pd.DataFrame( [ [0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00000000], [-0.00000000, -0.00000000, -0.00000000, -0.00000000, -0.00000000], [0.01859903, -0.09733263, 0.00232372, 0.00000000, -0.10228998], [-0.41819699, 0.00000000, -0.86088613, 2.30892279, 0.28210892]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_filter_signals(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(10) fn_inputs = { 'signal': pd.DataFrame( [[0, 0, 0], [-1, -1, -1], [1, 0, -1], [0, 0, 0], [1, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 1], [-1, 0, 0], [0, 0, 0]], dates, tickers), 'lookahead_days': 3 } fn_correct_outputs = OrderedDict([ ('filtered_signal', pd.DataFrame( [[0, 0, 0], [-1, -1, -1], [1, 0, 0], [0, 0, 0], [0, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 0], [-1, 0, 0], [0, 0, 0]], dates, tickers)) ]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_kstest(fn): tickers = generate_random_tickers(3) fn_inputs = { 'long_short_signal_returns': pd.DataFrame( { 'ticker': tickers * 5, 'signal_return': [0.12, -0.83, 0.37, 0.83, -0.34, 0.27, -0.68, 0.29, 0.69, 0.57, 0.39, 0.56, -0.97, -0.72, 0.26]})} fn_correct_outputs = OrderedDict([ ( 'ks_values', pd.Series( [0.29787827, 0.35221525, 0.63919407], tickers)), ( 'p_values', pd.Series( [0.69536353, 0.46493498, 0.01650327], tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_get_long_short(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'close': pd.DataFrame( [ [25.6788, 35.1392, 34.0527], [25.1884, 14.3453, 39.9373], [78.2803, 34.3854, 23.2932], [88.8725, 52.223, 34.4107]], dates, tickers), 'lookback_high': pd.DataFrame( [ [np.nan, np.nan, np.nan], [92.11310000, 91.05430000, 90.95720000], [35.4411, 34.1799, 34.0223], [92.11310000, 91.05430000, 90.95720000]], dates, tickers), 'lookback_low': pd.DataFrame( [ [np.nan, np.nan, np.nan], [34.1705, 92.453, 58.5107], [15.67180000, 12.34530000, 34.05270000], [27.18340000, 12.34530000, 23.29320000]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'long_short', pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 1, -1], [0, 0, 0]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_filter_signals(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(10) fn_inputs = { 'signal': pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 0, -1], [0, 0, 0], [1, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 1], [-1, 0, 0], [0, 0, 0]], dates, tickers), 'lookahead_days': 3} fn_correct_outputs = OrderedDict([ ( 'filtered_signal', pd.DataFrame( [ [0, 0, 0], [-1, -1, -1], [1, 0, 0], [0, 0, 0], [0, 0, 0], [0, 1, 0], [0, 0, 1], [0, -1, 0], [-1, 0, 0], [0, 0, 0]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_generate_returns(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'prices': pd.DataFrame( [ [35.4411, 34.1799, 34.0223], [92.1131, 91.0543, 90.9572], [57.9708, 57.7814, 58.1982], [34.1705, 92.453, 58.5107]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'returns', pd.DataFrame( [ [np.nan, np.nan, np.nan], [1.59904743, 1.66397210, 1.67345829], [-0.37065629, -0.36541822, -0.36015840], [-0.41055669, 0.60004777, 0.00536958]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_compute_log_returns(fn): tickers = generate_random_tickers(5) dates = pd.DatetimeIndex(['2008-08-31', '2008-09-30', '2008-10-31', '2008-11-30']) fn_inputs = { 'prices': pd.DataFrame( [ [21.05081048, 17.01384381, 10.98450376, 11.24809343, 12.96171273], [482.34539247, 35.20258059, 3516.54167823, 66.40531433, 13.50396048], [10.91893302, 17.90864387, 24.80126542, 12.48895419, 10.52435923], [11.54549538, 23.98146843, 24.97476306, 36.03196210, 14.30433232]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'log_returns', pd.DataFrame( [ [np.nan, np.nan, np.nan, np.nan, np.nan], [3.13172138, 0.72709204, 5.76874778, 1.77557845, 0.04098317], [-3.78816218, -0.67583590, -4.95433863, -1.67093250, -0.24929051], [0.05579709, 0.29199789, 0.00697116, 1.05956179, 0.30686995]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_calculate_dividend_weights(fn): tickers = generate_random_tickers(3) dates = generate_random_dates(4) fn_inputs = { 'dividends': pd.DataFrame( [ [0.0, 0.0, 0.0], [0.0, 0.0, 0.1], [0.0, 1.0, 0.3], [0.0, 0.2, 0.0]], dates, tickers)} fn_correct_outputs = OrderedDict([ ( 'dividend_weights', pd.DataFrame( [ [np.nan, np.nan, np.nan], [0.00000000, 0.00000000, 1.00000000], [0.00000000, 0.71428571, 0.28571429], [0.00000000, 0.75000000, 0.25000000]], dates, tickers))]) assert_output(fn, fn_inputs, fn_correct_outputs)
def test_find_outliers(fn): tickers = generate_random_tickers(7) dates = generate_random_dates(40) outlier_values = [{0.0: 1, 0.5: 3, 0.75: 13, 1.0: 12, 1.25: 8, 1.5: 3}] norm_values = [{ -1.0: 6, -0.5: 9, 0.0: 8, 0.5: 11, 1.0: 6 }, { -1.0: 4, -0.5: 13, 0.0: 9, 0.5: 11, 1.0: 3 }, { -1.0: 7, -0.5: 11, 0.0: 7, 0.5: 9, 1.0: 6 }, { -1.0: 5, -0.5: 9, 0.0: 11, 0.5: 10, 1.0: 5 }, { -1.0: 3, -0.5: 12, 0.0: 14, 0.5: 8, 1.0: 3 }, { -1.0: 5, -0.5: 13, 0.0: 7, 0.5: 9, 1.0: 6 }] # Build the values for the signal_return parameter signal_return_values = [] for values in outlier_values + norm_values: current_signal_return_values = [] for value, value_count in values.items(): current_signal_return_values.extend( [x for x in [value] * value_count]) signal_return_values.append(current_signal_return_values) signal_return_values = np.array(signal_return_values).T # Create signals that aren't the same for each date tile_value = [1, 1, 1, 0, 1] signal_values = np.tile( tile_value, int(np.product(signal_return_values.shape) / len(tile_value))).reshape( signal_return_values.shape) fn_inputs = { 'signal': pd.DataFrame(signal_values, dates, tickers), 'signal_return': pd.DataFrame(signal_return_values, dates, tickers), 'ks_threshold': 0.8 } fn_correct_outputs = OrderedDict([('outliers', [tickers[0]])]) assert_output(fn, fn_inputs, fn_correct_outputs)