因子预处理方法、多因子合成 - ChannelCMT/OFO GitHub Wiki
- 因子比较与筛选
- 常用的因子预处理方法-调整正负、去极值、行业市值中性化、标准化
- 多因子组合方法
from jaqs_fxdayu.data import DataView
import warnings
warnings.filterwarnings("ignore")
dataview_folder = '../Factor'
dv = DataView()
dv.load_dataview(dataview_folder)
dv.add_formula("momentum", "Return(close_adj, 20)", is_quarterly=False, add_data=True)Dataview loaded successfully.
| symbol | 000001.SZ | 000002.SZ | 000008.SZ | 000009.SZ | 000012.SZ | 000024.SZ | 000027.SZ | 000039.SZ | 000046.SZ | 000059.SZ | ... | 601998.SH | 603000.SH | 603160.SH | 603288.SH | 603699.SH | 603799.SH | 603833.SH | 603858.SH | 603885.SH | 603993.SH |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| trade_date | |||||||||||||||||||||
| 20140102 | -0.100735 | -0.085812 | -0.057592 | -0.006342 | -0.100442 | -0.051708 | -0.068143 | 0.012426 | -0.074534 | -0.089580 | ... | -0.065375 | 0.104574 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.084892 |
| 20140103 | -0.111690 | -0.102975 | -0.052910 | -0.040881 | -0.116740 | -0.078923 | -0.082474 | 0.048699 | -0.091097 | -0.111111 | ... | -0.075426 | 0.105497 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.091437 |
| 20140106 | -0.121896 | -0.137255 | -0.095643 | -0.059129 | -0.165380 | -0.111576 | -0.106164 | 0.011311 | -0.098121 | -0.134470 | ... | -0.085575 | 0.132137 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.123726 |
| 20140107 | -0.118271 | -0.138051 | -0.109342 | -0.060228 | -0.174342 | -0.122535 | -0.104991 | 0.039841 | -0.095745 | -0.139847 | ... | -0.088020 | 0.076545 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.118594 |
| 20140108 | -0.115124 | -0.144175 | -0.159346 | -0.063224 | -0.179235 | -0.160665 | -0.093103 | 0.066347 | -0.081023 | -0.156604 | ... | -0.085575 | 0.118630 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.127941 |
| 20140109 | -0.079439 | -0.126464 | -0.149474 | -0.050273 | -0.171525 | -0.145109 | -0.087873 | 0.109015 | -0.075107 | -0.143411 | ... | -0.081481 | 0.118595 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.157895 |
| 20140110 | -0.070755 | -0.134818 | -0.147799 | -0.048087 | -0.186020 | -0.158701 | -0.097002 | 0.088464 | -0.066667 | -0.141199 | ... | -0.077114 | 0.032493 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.150665 |
| 20140113 | -0.075697 | -0.152225 | -0.117032 | -0.038419 | -0.189342 | -0.207055 | -0.083333 | 0.074930 | -0.069414 | -0.193490 | ... | -0.062344 | 0.109781 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.113264 |
| 20140114 | -0.080784 | -0.142349 | -0.100105 | -0.044662 | -0.089120 | -0.187791 | -0.063063 | 0.123944 | -0.039560 | -0.177007 | ... | -0.062814 | 0.100159 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.094225 |
| 20140115 | -0.082547 | -0.133011 | -0.092827 | -0.015385 | -0.077739 | -0.182381 | -0.060000 | 0.153846 | -0.040000 | -0.162921 | ... | -0.078481 | 0.209846 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.095679 |
| 20140116 | -0.083856 | -0.129540 | -0.105152 | -0.045259 | -0.070071 | -0.178147 | -0.047016 | 0.177477 | -0.042411 | -0.148571 | ... | -0.081013 | 0.235830 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.081413 |
| 20140117 | -0.081600 | -0.130221 | -0.129474 | -0.064725 | -0.098927 | -0.168452 | -0.043956 | 0.160142 | -0.024943 | -0.172147 | ... | -0.084184 | 0.230942 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.101538 |
| 20140120 | -0.051217 | -0.091720 | -0.110410 | -0.049834 | -0.074390 | -0.141717 | -0.042831 | 0.183333 | -0.011494 | -0.152174 | ... | 0.002793 | 0.216201 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.097638 |
| 20140121 | -0.049372 | -0.073171 | -0.117284 | -0.044199 | -0.053528 | -0.131671 | -0.037175 | 0.226483 | 0.000000 | -0.139721 | ... | -0.042328 | 0.280011 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.090625 |
| 20140122 | 0.007692 | -0.026820 | -0.109645 | -0.011086 | -0.008750 | -0.061973 | -0.024074 | 0.222865 | 0.013793 | -0.109091 | ... | -0.034211 | 0.282502 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.070203 |
| 20140123 | -0.001706 | -0.053097 | -0.085263 | -0.017544 | -0.009828 | -0.081773 | -0.025878 | 0.137441 | -0.037445 | -0.115152 | ... | -0.037037 | 0.199393 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.075000 |
| 20140124 | 0.018341 | -0.001282 | -0.038803 | 0.016484 | 0.015152 | -0.039098 | -0.018587 | 0.117726 | -0.028139 | -0.103659 | ... | -0.021448 | 0.246027 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.063665 |
| 20140127 | -0.030638 | -0.047323 | -0.053512 | 0.007600 | -0.017478 | -0.086402 | -0.034926 | 0.056916 | -0.039216 | -0.117172 | ... | -0.042216 | 0.162772 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.077399 |
| 20140128 | -0.020443 | -0.027990 | -0.049889 | 0.006515 | 0.003727 | -0.064851 | -0.033088 | 0.059133 | -0.024609 | -0.112903 | ... | -0.023810 | 0.114058 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.074419 |
| 20140129 | -0.056327 | -0.061021 | -0.048889 | 0.000000 | -0.009828 | -0.093840 | -0.043636 | 0.057873 | -0.020045 | -0.116232 | ... | -0.038760 | 0.220854 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.078582 |
| 20140130 | -0.067866 | -0.076345 | -0.047778 | -0.003191 | -0.025767 | -0.099318 | -0.047532 | 0.017442 | -0.040268 | -0.146586 | ... | -0.064767 | 0.181957 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.073899 |
| 20140207 | -0.046102 | -0.054847 | -0.027933 | 0.045902 | 0.007481 | -0.082219 | -0.018727 | 0.015267 | -0.013667 | -0.127083 | ... | -0.042105 | 0.205879 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | -0.052716 |
| 20140210 | -0.008569 | 0.004011 | 0.038778 | 0.160970 | 0.097754 | -0.010989 | 0.021073 | 0.046711 | 0.013889 | -0.054705 | ... | 0.016043 | 0.212243 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.013289 |
| 20140211 | 0.031814 | 0.016151 | 0.052443 | 0.171271 | 0.095618 | 0.005350 | 0.028846 | 0.007024 | 0.032941 | -0.020045 | ... | 0.120643 | 0.125982 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.014950 |
| 20140212 | 0.015306 | 0.013477 | 0.089915 | 0.203600 | 0.125166 | 0.033003 | 0.039924 | 0.050032 | 0.011601 | -0.008949 | ... | 0.208556 | 0.080123 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.032040 |
| 20140213 | 0.033841 | -0.006702 | 0.092822 | 0.250863 | 0.115020 | -0.004338 | 0.040462 | 0.011972 | 0.039443 | 0.009050 | ... | 0.330645 | 0.055781 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.055556 |
| 20140214 | 0.021997 | -0.001355 | 0.107011 | 0.259472 | 0.149584 | 0.003861 | 0.066406 | 0.048765 | 0.041475 | 0.018018 | ... | 0.309973 | 0.108285 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.071304 |
| 20140217 | 0.031897 | 0.017956 | 0.108876 | 0.252283 | 0.181818 | 0.042750 | 0.075435 | 0.053420 | 0.062937 | 0.017937 | ... | 0.279255 | 0.054795 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.043697 |
| 20140218 | -0.005119 | 0.005533 | 0.080796 | 0.299886 | 0.072427 | 0.012622 | 0.053846 | -0.013158 | 0.054920 | 0.002217 | ... | 0.257373 | 0.038174 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.026846 |
| 20140219 | 0.028278 | 0.027894 | 0.090698 | 0.252232 | 0.067688 | 0.041351 | 0.061896 | -0.036137 | 0.064815 | 0.038031 | ... | 0.417582 | -0.021494 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | 0.044369 |
| ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
| 20171120 | 0.273458 | 0.114539 | 0.003382 | -0.106933 | 0.073761 | NaN | -0.051084 | 0.062434 | -0.076167 | -0.037791 | ... | -0.012598 | -0.095861 | 0.072122 | 0.074252 | 0.116348 | -0.059919 | 0.034426 | -0.205057 | 0.033520 | -0.126404 |
| 20171121 | 0.268657 | 0.177407 | -0.003378 | -0.105077 | 0.090686 | NaN | -0.046512 | 0.098352 | -0.075887 | -0.033203 | ... | -0.010955 | -0.081181 | 0.053217 | 0.093856 | 0.123353 | -0.060377 | 0.044169 | -0.205060 | 0.037089 | -0.153740 |
| 20171122 | 0.339840 | 0.184834 | -0.004484 | -0.104460 | 0.132927 | NaN | -0.049536 | 0.073584 | -0.069767 | -0.027079 | ... | 0.020408 | -0.086637 | 0.068177 | 0.108031 | 0.118350 | -0.030201 | -0.048057 | -0.236038 | 0.048184 | -0.141667 |
| 20171123 | 0.283542 | 0.136942 | -0.065539 | -0.129070 | 0.092457 | NaN | -0.065015 | 0.027455 | -0.045455 | -0.075914 | ... | 0.001585 | -0.109250 | 0.163450 | -0.007020 | 0.080305 | -0.052761 | -0.055597 | -0.243516 | -0.007785 | -0.148098 |
| 20171124 | 0.259516 | 0.156364 | -0.071506 | -0.096085 | 0.141983 | NaN | -0.057632 | 0.030655 | -0.025894 | -0.065789 | ... | 0.000000 | -0.109810 | 0.208555 | -0.009451 | 0.071470 | -0.041119 | -0.065847 | -0.248355 | 0.014144 | -0.078404 |
| 20171127 | 0.205017 | 0.087647 | -0.108719 | -0.076638 | 0.124069 | NaN | -0.050394 | 0.029714 | -0.015365 | -0.076493 | ... | 0.019293 | -0.073933 | 0.088818 | -0.035897 | 0.062900 | -0.104015 | -0.069122 | -0.247189 | 0.000000 | -0.045190 |
| 20171128 | 0.187175 | 0.061809 | -0.091503 | -0.066258 | 0.115942 | NaN | -0.051643 | 0.003638 | -0.012870 | -0.047214 | ... | 0.009646 | -0.078550 | 0.121669 | -0.012093 | 0.036179 | -0.119827 | -0.038106 | -0.248399 | -0.038624 | -0.047420 |
| 20171129 | 0.212281 | 0.160206 | -0.088428 | -0.045735 | 0.144242 | NaN | -0.053208 | 0.012671 | 0.003851 | -0.018112 | ... | 0.011364 | -0.088722 | 0.004145 | -0.011621 | 0.041572 | -0.157026 | -0.074108 | -0.242310 | -0.037657 | -0.033103 |
| 20171130 | 0.159445 | 0.060102 | -0.075245 | -0.045226 | 0.086288 | NaN | -0.036220 | -0.010660 | -0.014175 | -0.006843 | ... | 0.039216 | -0.073933 | -0.014040 | -0.031875 | 0.036261 | -0.121046 | -0.065931 | -0.243670 | -0.048193 | 0.008559 |
| 20171201 | 0.141352 | 0.090103 | -0.070346 | -0.024453 | 0.084524 | NaN | -0.036335 | 0.015377 | -0.029601 | 0.006958 | ... | 0.042414 | -0.051938 | -0.004686 | -0.035088 | 0.024650 | -0.111065 | -0.049722 | -0.225288 | -0.062633 | 0.028986 |
| 20171204 | 0.179078 | 0.123452 | -0.071817 | -0.052363 | 0.102259 | NaN | -0.045741 | -0.011946 | -0.050761 | -0.018465 | ... | 0.055738 | -0.061303 | -0.090380 | -0.077368 | -0.012379 | -0.059388 | -0.102398 | -0.269209 | -0.005780 | 0.014514 |
| 20171205 | 0.115772 | 0.114183 | -0.062569 | -0.085025 | 0.046484 | NaN | -0.065831 | -0.050197 | -0.053571 | -0.087324 | ... | 0.055105 | -0.111872 | -0.110107 | -0.103594 | -0.028479 | -0.093309 | -0.099602 | -0.273266 | -0.013889 | -0.042029 |
| 20171206 | 0.078318 | 0.113242 | -0.051339 | -0.093284 | 0.058683 | NaN | -0.070203 | -0.030378 | -0.043871 | -0.081056 | ... | 0.033926 | -0.120986 | -0.065739 | -0.095816 | -0.018095 | -0.108374 | -0.086371 | -0.269365 | -0.018248 | -0.047688 |
| 20171207 | 0.040552 | 0.082400 | -0.068462 | -0.088639 | 0.034606 | NaN | -0.067083 | -0.037365 | -0.032258 | -0.078154 | ... | 0.035599 | -0.124720 | -0.112329 | -0.108496 | -0.022608 | -0.126953 | -0.139220 | -0.269252 | -0.027201 | -0.072674 |
| 20171208 | 0.064228 | 0.083969 | -0.062153 | -0.090683 | 0.060680 | NaN | -0.054859 | -0.071258 | -0.029525 | -0.090226 | ... | 0.037582 | -0.113619 | -0.047843 | -0.096609 | -0.001064 | -0.117551 | -0.119264 | -0.247828 | -0.034261 | -0.023669 |
| 20171211 | 0.046512 | 0.106061 | -0.034792 | -0.069913 | -0.030905 | NaN | -0.045455 | -0.033157 | -0.032092 | -0.080299 | ... | 0.029126 | -0.097358 | -0.085663 | -0.047806 | 0.004787 | -0.034339 | -0.098188 | -0.246069 | -0.013591 | 0.010432 |
| 20171212 | 0.005405 | 0.037544 | -0.061590 | -0.052564 | -0.029834 | NaN | -0.047393 | -0.018031 | -0.020699 | -0.082464 | ... | 0.024631 | -0.069124 | -0.069708 | -0.021772 | 0.012299 | 0.085595 | -0.077405 | -0.242025 | -0.022615 | 0.014793 |
| 20171213 | 0.017829 | 0.052743 | -0.084895 | -0.045103 | -0.003386 | NaN | -0.048742 | 0.006317 | -0.019506 | -0.060136 | ... | 0.027915 | -0.090701 | 0.006163 | 0.007490 | 0.023555 | 0.132361 | -0.070957 | -0.217561 | 0.013204 | 0.056250 |
| 20171214 | -0.007634 | 0.040428 | -0.073579 | -0.063226 | -0.010169 | NaN | -0.039683 | 0.020088 | 0.006729 | -0.071636 | ... | 0.028286 | -0.101592 | 0.003662 | -0.015745 | 0.015508 | 0.143427 | -0.044972 | -0.213025 | 0.013841 | 0.108766 |
| 20171215 | -0.034901 | 0.007266 | -0.056561 | -0.025401 | -0.022196 | NaN | -0.016103 | 0.050813 | 0.069307 | -0.062937 | ... | -0.001631 | -0.070356 | 0.004673 | 0.005391 | 0.030769 | 0.113165 | 0.008809 | -0.144016 | 0.009729 | 0.097682 |
| 20171218 | -0.105263 | -0.015225 | -0.053933 | -0.046053 | -0.074324 | NaN | -0.013051 | 0.019422 | 0.007979 | -0.060423 | ... | -0.022329 | -0.067470 | -0.048306 | -0.021110 | -0.004276 | -0.013904 | -0.005952 | -0.095776 | -0.035811 | 0.059486 |
| 20171219 | -0.080969 | -0.082730 | -0.038418 | -0.044855 | -0.074157 | NaN | -0.013008 | 0.068248 | 0.010596 | -0.045455 | ... | -0.020570 | -0.059438 | -0.024976 | -0.029272 | -0.004264 | -0.024461 | 0.019060 | -0.085810 | -0.021592 | 0.075286 |
| 20171220 | -0.121854 | -0.120000 | -0.040541 | -0.064220 | -0.127018 | NaN | -0.014658 | 0.077909 | -0.006579 | -0.076541 | ... | -0.047692 | -0.072347 | -0.087828 | -0.005519 | 0.010689 | -0.051383 | 0.049189 | -0.111150 | -0.032645 | 0.035599 |
| 20171221 | -0.056446 | -0.050479 | -0.041855 | -0.052069 | -0.087973 | NaN | 0.003311 | 0.137718 | -0.033462 | -0.051724 | ... | -0.022152 | -0.072772 | -0.171216 | 0.063826 | 0.026587 | -0.032383 | 0.079497 | -0.082461 | 0.052782 | 0.036683 |
| 20171222 | -0.071429 | -0.061321 | -0.007001 | -0.051181 | -0.114684 | NaN | 0.016529 | 0.113846 | -0.050633 | -0.048290 | ... | -0.022117 | -0.073191 | -0.221484 | 0.063208 | 0.027115 | -0.057867 | 0.075556 | -0.080550 | 0.040446 | -0.038806 |
| 20171225 | -0.048816 | -0.036485 | 0.028986 | -0.040161 | -0.105960 | NaN | 0.004975 | 0.143232 | -0.024707 | -0.073737 | ... | -0.028391 | -0.113580 | -0.171475 | 0.099632 | 0.031233 | 0.048498 | 0.099036 | -0.077581 | 0.058182 | -0.016794 |
| 20171226 | -0.002920 | -0.008130 | 0.049161 | -0.055191 | -0.109307 | NaN | 0.001650 | 0.144485 | -0.027379 | -0.074331 | ... | -0.007962 | -0.113115 | -0.176979 | 0.042646 | 0.044190 | -0.026917 | 0.054350 | -0.059463 | 0.086194 | -0.060029 |
| 20171227 | -0.038350 | -0.089592 | 0.044311 | -0.067358 | -0.136653 | NaN | -0.001653 | 0.079580 | -0.046036 | -0.092233 | ... | -0.004815 | -0.115512 | -0.158704 | 0.070140 | 0.037726 | -0.011718 | 0.065846 | -0.062373 | 0.057971 | -0.064194 |
| 20171228 | -0.012706 | -0.016656 | 0.021226 | -0.055263 | -0.108814 | NaN | -0.014706 | 0.124166 | -0.026144 | -0.076772 | ... | -0.028302 | -0.111934 | -0.148128 | 0.089162 | 0.032805 | -0.042601 | 0.055350 | -0.074094 | 0.102755 | -0.008487 |
| 20171229 | 0.023077 | 0.010739 | 0.018626 | -0.046174 | -0.072448 | NaN | -0.006557 | 0.153458 | -0.010610 | -0.059230 | ... | -0.029734 | -0.114473 | -0.122397 | 0.099081 | 0.051394 | -0.069258 | 0.045153 | -0.066275 | 0.160972 | -0.030986 |
977 rows × 488 columns
import numpy as np
def mask_index_member():
df_index_member = dv.get_ts('index_member')
mask_index_member = ~(df_index_member >0) #定义信号过滤条件-非指数成分
return mask_index_member
def limit_up_down():
# 定义可买卖条件——未停牌、未涨跌停
trade_status = dv.get_ts('trade_status')
mask_sus = trade_status == 0
# 涨停
dv.add_formula('up_limit', '(close - Delay(close, 1)) / Delay(close, 1) > 0.095', is_quarterly=False, add_data=True)
# 跌停
dv.add_formula('down_limit', '(close - Delay(close, 1)) / Delay(close, 1) < -0.095', is_quarterly=False, add_data=True)
can_enter = np.logical_and(dv.get_ts('up_limit') < 1, ~mask_sus) # 未涨停未停牌
can_exit = np.logical_and(dv.get_ts('down_limit') < 1, ~mask_sus) # 未跌停未停牌
return can_enter,can_exit
mask = mask_index_member()
can_enter,can_exit = limit_up_down()接下来，我们对pb、pe、ps、float_mv、momentum五个因子进行比较、筛选
from jaqs_fxdayu.research.signaldigger import multi_factor
ic = dict()
factors_dict = {signal:dv.get_ts(signal) for signal in ["pb","pe","ps","float_mv","momentum"]}
for period in [5, 15, 30]:
ic[period]=multi_factor.get_factors_ic_df(factors_dict,
price=dv.get_ts("close_adj"),
high=dv.get_ts("high_adj"), # 可为空
low=dv.get_ts("low_adj"),# 可为空
n_quantiles=5,# quantile分类数
mask=mask,# 过滤条件
can_enter=can_enter,# 是否能进场
can_exit=can_exit,# 是否能出场
period=period,# 持有期
benchmark_price=dv.data_benchmark, # 基准价格 可不传入，持有期收益（return）计算为绝对收益
commission = 0.0008,
)Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
Nan Data Count (should be zero) : 0; Percentage of effective data: 58%
import pandas as pd
ic_mean_table = pd.DataFrame(data=np.nan,columns=[5,15,30],index=["pb","pe","ps","float_mv","momentum"])
ic_std_table = pd.DataFrame(data=np.nan,columns=[5,15,30],index=["pb","pe","ps","float_mv","momentum"])
ir_table = pd.DataFrame(data=np.nan,columns=[5,15,30],index=["pb","pe","ps","float_mv","momentum"])
for signal in ["pb","pe","ps","float_mv","momentum"]:
for period in [5, 15, 30]:
ic_mean_table.loc[signal,period]=ic[period][signal].mean()
ic_std_table.loc[signal,period]=ic[period][signal].std()
ir_table.loc[signal,period]=ic[period][signal].mean()/ic[period][signal].std()
print(ic_mean_table)
print(ic_std_table)
print(ir_table) 5 15 30
pb -0.039948 -0.069184 -0.106428
pe -0.038036 -0.065607 -0.098353
ps -0.032231 -0.057777 -0.087181
float_mv 0.006833 0.021287 0.044382
momentum -0.041551 -0.053251 -0.047145
5 15 30
pb 0.231587 0.259397 0.245520
pe 0.210134 0.220244 0.210795
ps 0.176345 0.193792 0.188749
float_mv 0.222908 0.229546 0.229144
momentum 0.207719 0.215057 0.209887
5 15 30
pb -0.172496 -0.266712 -0.433481
pe -0.181008 -0.297881 -0.466578
ps -0.182774 -0.298140 -0.461889
float_mv 0.030655 0.092735 0.193688
momentum -0.200034 -0.247614 -0.224622
可视化比较
%matplotlib inline
ic_mean_table.plot(kind="barh",xerr=ic_std_table,figsize=(15,5))<matplotlib.axes._subplots.AxesSubplot at 0x18b5268e710>
- IC_IR:方差标准化后的ic均值
- 一般而言，认为|IC_IR|>0.6,因子的稳定性合格
%matplotlib inline
ir_table.plot(kind="barh",figsize=(15,5))<matplotlib.axes._subplots.AxesSubplot at 0x18b4fb3fd68>
保留momentum、ps、pe、pb 进一步处理并尝试构建组合因子
- 根据之前的分析，这几个因子在几个持有期下与股票收益的关系（ic）都是负的，先统一调整成正相关关系
- 去极值
- 标准化 -- z-score、rank
- 行业市值中性化
from jaqs_fxdayu.research.signaldigger import process
factor_dict = dict()
index_member = dv.get_ts("index_member")
for name in ["pb","pe","ps","momentum"]:
signal = -1*dv.get_ts(name) # 调整符号
process.winsorize(factor_df=signal,alpha=0.05,index_member=index_member)#去极值
signal = process.standardize(signal,index_member) #z-score标准化 保留排序信息和分布信息
# signal = process.rank_standardize(signal,index_member) #因子在截面排序并归一化到0-1(只保留排序信息)
# # 行业市值中性化
# signal = process.neutralize(signal,
# group=dv.get_ts("sw1"),# 行业分类标准
# float_mv = dv.get_ts("float_mv"), #流通市值 可为None 则不进行市值中性化
# index_member=index_member,# 是否只处理时只考虑指数成份股
# )
factor_dict[name] = signal对筛选后的因子进行组合，一般有以下常规处理：
- 因子间存在较强同质性时，先使用施密特正交化方法对因子做正交化处理，用得到的正交化残差作为因子（也可以不使用，正交化会破坏因子的经济学逻辑，并剔除一些信息）
- 因子组合加权，常规的方法有：等权重、以某个时间窗口的滚动平均ic为权重、以某个时间窗口的滚动ic_ir为权重、最大化上个持有期的ic_ir为目标处理权重、最大化上个持有期的ic为目标处理权重
- 注:因为计算IC需要用到下一期股票收益,因此在动态加权方法里，实际上使用的是前一期及更早的IC值(向前推移了holding_period)计算当期的权重
# 因子间存在较强同质性时，使用施密特正交化方法对因子做正交化处理，用得到的正交化残差作为因子
new_factors = multi_factor.orthogonalize(factors_dict=factor_dict,
standardize_type="rank",#输入因子标准化方法，有"rank"（排序标准化）,"z_score"(z-score标准化)两种（"rank"/"z_score"）
winsorization=False,#是否对输入因子去极值
index_member=index_member) # 是否只处理指数成分股new_factors{'momentum': symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140102 0.715719 0.511706 NaN 0.354515 0.290970 0.709030
20140103 0.668896 0.491639 NaN 0.351171 0.260870 0.655518
20140106 0.722408 0.488294 NaN 0.354515 0.230769 0.645485
20140107 0.725753 0.488294 NaN 0.384615 0.190635 0.605351
20140108 0.745819 0.498328 NaN 0.367893 0.200669 0.471572
20140109 0.816054 0.531773 NaN 0.371237 0.234114 0.511706
20140110 0.856187 0.505017 NaN 0.394649 0.227425 0.474916
20140113 0.842809 0.484950 NaN 0.444816 0.220736 0.334448
20140114 0.812709 0.418060 NaN 0.290970 0.414716 0.341137
20140115 0.792642 0.451505 NaN 0.397993 0.464883 0.337793
20140116 0.792642 0.428094 NaN 0.317726 0.481605 0.294314
20140117 0.822742 0.461538 NaN 0.257525 0.414716 0.394649
... ... ... ... ... ... ...
20171221 0.759197 0.351171 0.792642 NaN NaN NaN
20171222 0.759197 0.351171 0.795987 NaN NaN NaN
20171225 0.698997 0.301003 0.812709 NaN NaN NaN
20171226 0.779264 0.334448 0.795987 NaN NaN NaN
20171227 0.250836 0.715719 0.197324 NaN NaN NaN
20171228 0.799331 0.304348 0.809365 NaN NaN NaN
20171229 0.832776 0.321070 0.779264 NaN NaN NaN
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140102 NaN 0.668896 0.622074 NaN ... 0.812709
20140103 NaN 0.852843 0.541806 NaN ... 0.795987
20140106 NaN 0.816054 0.628763 NaN ... 0.822742
20140107 NaN 0.852843 0.658863 NaN ... 0.809365
20140108 NaN 0.892977 0.695652 NaN ... 0.829431
20140109 NaN 0.892977 0.719064 NaN ... 0.806020
20140110 NaN 0.866221 0.765886 NaN ... 0.819398
20140113 NaN 0.846154 0.769231 NaN ... 0.856187
20140114 NaN 0.862876 0.769231 NaN ... 0.829431
20140115 NaN 0.903010 0.755853 NaN ... 0.795987
20140116 NaN 0.926421 0.749164 NaN ... 0.789298
20171219 NaN NaN NaN NaN ... 0.257525
20171220 NaN NaN NaN NaN ... 0.210702
20171221 NaN NaN NaN NaN ... 0.816054
20171222 NaN NaN NaN NaN ... 0.816054
20171225 NaN NaN NaN NaN ... 0.866221
20171226 NaN NaN NaN NaN ... 0.829431
20171227 NaN NaN NaN NaN ... 0.147157
20171228 NaN NaN NaN NaN ... 0.826087
20171229 NaN NaN NaN NaN ... 0.812709
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140102 0.959866 NaN NaN NaN NaN NaN
20140103 0.949833 NaN NaN NaN NaN NaN
20140106 0.953177 NaN NaN NaN NaN NaN
20140107 0.916388 NaN NaN NaN NaN NaN
20140108 0.913043 NaN NaN NaN NaN NaN
20140109 0.946488 NaN NaN NaN NaN NaN
20140110 0.946488 NaN NaN NaN NaN NaN
20140113 0.976589 NaN NaN NaN NaN NaN
... ... ... ... ... ... ...
20171212 NaN 0.973244 NaN NaN 0.986622 0.916388
20171213 NaN 0.976589 NaN NaN 0.986622 0.923077
20171214 NaN 0.976589 NaN NaN 0.986622 0.939799
20171215 NaN 0.976589 NaN NaN 0.986622 0.953177
20171218 NaN 0.973244 NaN NaN 0.979933 0.953177
20171219 NaN 0.979933 NaN NaN 0.976589 0.953177
20171220 NaN 0.963211 NaN NaN 0.979933 0.953177
20171221 NaN 0.023411 NaN NaN 0.026756 0.053512
20171222 NaN 0.026756 NaN NaN 0.023411 0.053512
20171225 NaN 0.013378 NaN NaN 0.030100 0.086957
20171226 NaN 0.023411 NaN NaN 0.026756 0.056856
20171227 NaN 0.976589 NaN NaN 0.973244 0.936455
20171228 NaN 0.016722 NaN NaN 0.026756 0.063545
20171229 NaN 0.016722 NaN NaN 0.026756 0.063545
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140102 NaN NaN 0.605351
20140103 NaN NaN 0.548495
20140106 NaN NaN 0.521739
20140107 NaN NaN 0.565217
20140108 NaN NaN 0.511706
20140109 NaN NaN 0.438127
20140110 NaN NaN 0.521739
20140113 NaN NaN 0.648829
20140114 NaN NaN 0.655518
20140115 NaN NaN 0.591973
20140116 NaN NaN 0.632107
20140117 NaN NaN 0.632107
20140120 NaN NaN 0.638796
20140121 NaN NaN 0.759197
20140122 NaN NaN 0.782609
20140123 NaN NaN 0.752508
20140124 NaN NaN 0.709030
20140127 NaN NaN 0.354515
... ... ... ...
20171225 0.287625 NaN 0.953177
20171226 0.384615 NaN 0.963211
20171227 0.615385 NaN 0.033445
20171228 0.274247 NaN 0.969900
20171229 0.337793 NaN 0.969900
[977 rows x 488 columns],
用正交化前的因子，分别进行等权、以某个时间窗口的滚动平均ic为权重、以某个时间窗口的滚动ic_ir为权重、最大化上个持有期的ic_ir为目标处理权重、最大化上个持有期的ic为目标处理权重的加权组合方式，然后测试组合因子表现
# rollback_period代表滚动窗口所用到的天数,即用前多少期的数据来计算现阶段的因子权重。 通常建议设置时间在半年以上,可以获得相对稳定的预期结果
# 多因子组合-动态加权参数配置
props = {
'price':dv.get_ts("close_adj"),
'high':dv.get_ts("high_adj"), # 可为空
'low':dv.get_ts("low_adj"),# 可为空
'ret_type': 'return',#可选参数还有upside_ret/downside_ret 则组合因子将以优化潜在上行、下行空间为目标
'benchmark_price': dv.data_benchmark, # 为空计算的是绝对收益 不为空计算相对收益
'period': 30, # 30天的持有期
'mask': mask,
'can_enter': can_enter,
'can_exit': can_exit,
'forward': True,
'commission': 0.0008,
"covariance_type": "shrink", # 协方差矩阵估算方法 还可以为"simple"
"rollback_period": 120} # 滚动窗口天数comb_factors = dict()
for method in ["equal_weight","ic_weight","ir_weight","max_IR","max_IC"]:
comb_factors[method] = multi_factor.combine_factors(factor_dict,
standardize_type="rank",
winsorization=False,
weighted_method=method,
props=props)
print(method)
print(comb_factors[method].dropna(how="all").head())equal_weight
symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140102 0.762542 0.819398 NaN 0.143813 0.397993 0.628763
20140103 0.745819 0.822742 NaN 0.187291 0.401338 0.648829
20140106 0.712375 0.842809 NaN 0.190635 0.434783 0.678930
20140107 0.705686 0.849498 NaN 0.190635 0.458194 0.702341
20140108 0.678930 0.842809 NaN 0.204013 0.444816 0.769231
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140102 NaN 0.464883 0.364548 NaN ... 0.765886
20140103 NaN 0.431438 0.377926 NaN ... 0.749164
20140106 NaN 0.441472 0.351171 NaN ... 0.715719
20140107 NaN 0.404682 0.347826 NaN ... 0.719064
20140108 NaN 0.408027 0.324415 NaN ... 0.712375
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140102 0.0 NaN NaN NaN NaN NaN
20140103 0.0 NaN NaN NaN NaN NaN
20140106 0.0 NaN NaN NaN NaN NaN
20140107 0.0 NaN NaN NaN NaN NaN
20140108 0.0 NaN NaN NaN NaN NaN
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140102 NaN NaN 0.344482
20140103 NaN NaN 0.347826
20140106 NaN NaN 0.364548
20140107 NaN NaN 0.351171
20140108 NaN NaN 0.367893
[5 rows x 488 columns]
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
ic_weight
symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140812 0.775920 0.826087 NaN 0.297659 0.655518 0.839465
20140813 0.755853 0.789298 NaN 0.311037 0.642140 0.869565
20140814 0.762542 0.799331 NaN 0.307692 0.652174 0.849498
20140815 0.762542 0.852843 NaN 0.153846 0.595318 0.906355
20140818 0.765886 0.913043 NaN 0.083612 0.585284 0.929766
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140812 0.829431 0.698997 NaN NaN ... 0.836120
20140813 0.775920 0.705686 NaN NaN ... 0.856187
20140814 0.846154 0.719064 NaN NaN ... 0.852843
20140815 0.792642 0.719064 NaN NaN ... 0.826087
20140818 0.749164 0.719064 NaN NaN ... 0.836120
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140812 0.070234 NaN NaN 0.207358 NaN NaN
20140813 0.083612 NaN NaN 0.197324 NaN NaN
20140814 0.080268 NaN NaN 0.204013 NaN NaN
20140815 0.090301 NaN NaN 0.183946 NaN NaN
20140818 0.050167 NaN NaN 0.163880 NaN NaN
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140812 NaN NaN 0.153846
20140813 NaN NaN 0.157191
20140814 NaN NaN 0.311037
20140815 NaN NaN 0.301003
20140818 NaN NaN 0.270903
[5 rows x 488 columns]
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
ir_weight
symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140812 0.769231 0.859532 NaN 0.311037 0.658863 0.842809
20140813 0.732441 0.819398 NaN 0.331104 0.655518 0.876254
20140814 0.732441 0.819398 NaN 0.331104 0.665552 0.849498
20140815 0.739130 0.872910 NaN 0.170569 0.605351 0.919732
20140818 0.732441 0.933110 NaN 0.073579 0.595318 0.936455
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140812 0.829431 0.732441 NaN NaN ... 0.816054
20140813 0.769231 0.739130 NaN NaN ... 0.839465
20140814 0.839465 0.749164 NaN NaN ... 0.826087
20140815 0.785953 0.745819 NaN NaN ... 0.816054
20140818 0.715719 0.752508 NaN NaN ... 0.819398
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140812 0.080268 NaN NaN 0.234114 NaN NaN
20140813 0.093645 NaN NaN 0.247492 NaN NaN
20140814 0.090301 NaN NaN 0.240803 NaN NaN
20140815 0.110368 NaN NaN 0.227425 NaN NaN
20140818 0.080268 NaN NaN 0.207358 NaN NaN
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140812 NaN NaN 0.140468
20140813 NaN NaN 0.147157
20140814 NaN NaN 0.321070
20140815 NaN NaN 0.307692
20140818 NaN NaN 0.247492
[5 rows x 488 columns]
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
max_IR
symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140813 0.331104 0.468227 NaN 0.682274 0.230769 0.645485
20140814 0.374582 0.478261 NaN 0.682274 0.237458 0.561873
20140815 0.414716 0.652174 NaN 0.387960 0.160535 0.705686
20140818 0.408027 0.739130 NaN 0.173913 0.163880 0.765886
20140819 0.501672 0.765886 NaN 0.120401 0.210702 0.826087
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140813 0.625418 0.779264 NaN NaN ... 0.461538
20140814 0.755853 0.795987 NaN NaN ... 0.458194
20140815 0.678930 0.799331 NaN NaN ... 0.451505
20140818 0.591973 0.799331 NaN NaN ... 0.461538
20140819 0.662207 0.792642 NaN NaN ... 0.481605
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140813 0.371237 NaN NaN 0.147157 NaN NaN
20140814 0.371237 NaN NaN 0.170569 NaN NaN
20140815 0.391304 NaN NaN 0.130435 NaN NaN
20140818 0.280936 NaN NaN 0.100334 NaN NaN
20140819 0.026756 NaN NaN 0.183946 NaN NaN
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140813 NaN NaN 0.100334
20140814 NaN NaN 0.418060
20140815 NaN NaN 0.341137
20140818 NaN NaN 0.224080
20140819 NaN NaN 0.257525
[5 rows x 488 columns]
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
max_IC
symbol 000001.SZ 000002.SZ 000008.SZ 000009.SZ 000012.SZ 000024.SZ \
trade_date
20140221 0.030100 0.324415 NaN 0.903010 0.494983 0.127090
20140224 0.020067 0.163880 NaN 0.956522 0.568562 0.076923
20140225 0.193980 0.672241 NaN 0.451505 0.428094 0.521739
20140226 0.341137 0.903010 NaN 0.120401 0.471572 0.849498
20140227 0.471572 0.799331 NaN 0.170569 0.571906 0.849498
symbol 000027.SZ 000039.SZ 000046.SZ 000059.SZ ... 601998.SH \
trade_date ...
20140221 NaN 0.464883 0.511706 NaN ... 0.387960
20140224 NaN 0.464883 0.401338 NaN ... 0.244147
20140225 NaN 0.749164 0.240803 NaN ... 0.317726
20140226 NaN 0.983278 0.280936 NaN ... 0.013378
20140227 NaN 0.866221 0.180602 NaN ... 0.160535
symbol 603000.SH 603160.SH 603288.SH 603699.SH 603799.SH 603833.SH \
trade_date
20140221 0.311037 NaN NaN NaN NaN NaN
20140224 0.571906 NaN NaN NaN NaN NaN
20140225 0.040134 NaN NaN NaN NaN NaN
20140226 0.461538 NaN NaN NaN NaN NaN
20140227 0.193980 NaN NaN NaN NaN NaN
symbol 603858.SH 603885.SH 603993.SH
trade_date
20140221 NaN NaN 0.381271
20140224 NaN NaN 0.451505
20140225 NaN NaN 0.170569
20140226 NaN NaN 0.294314
20140227 NaN NaN 0.254181
[5 rows x 488 columns]
period = 30
ic_30 = multi_factor.get_factors_ic_df(comb_factors,
price=dv.get_ts("close_adj"),
high=dv.get_ts("high_adj"), # 可为空
low=dv.get_ts("low_adj"),# 可为空
n_quantiles=5,# quantile分类数
mask=mask,# 过滤条件
can_enter=can_enter,# 是否能进场
can_exit=can_exit,# 是否能出场
period=period,# 持有期
benchmark_price=dv.data_benchmark, # 基准价格 可不传入，持有期收益（return）计算为绝对收益
commission = 0.0008,
)
ic_30 = pd.concat([ic_30,-1*ic[30].drop("float_mv",axis=1)],axis=1)
ic_30.head()Nan Data Count (should be zero) : 0; Percentage of effective data: 57%
Nan Data Count (should be zero) : 0; Percentage of effective data: 48%
Nan Data Count (should be zero) : 0; Percentage of effective data: 48%
Nan Data Count (should be zero) : 0; Percentage of effective data: 48%
Nan Data Count (should be zero) : 0; Percentage of effective data: 56%
| equal_weight | ic_weight | ir_weight | max_IR | max_IC | pb | pe | ps | momentum | |
|---|---|---|---|---|---|---|---|---|---|
| trade_date | |||||||||
| 20140102 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| 20140103 | -0.046945 | NaN | NaN | NaN | NaN | -0.053375 | -0.018784 | -0.004749 | -0.050374 |
| 20140106 | -0.075316 | NaN | NaN | NaN | NaN | -0.085169 | -0.053065 | -0.018863 | -0.065761 |
| 20140107 | 0.027397 | NaN | NaN | NaN | NaN | 0.026080 | 0.023327 | 0.056947 | 0.013767 |
| 20140108 | 0.131549 | NaN | NaN | NaN | NaN | 0.084499 | 0.081695 | 0.158560 | 0.132101 |
ic_30_mean = dict()
ic_30_std = dict()
ir_30 = dict()
for name in ic_30.columns:
ic_30_mean[name]=ic_30[name].loc[20140901:].mean()
ic_30_std[name]=ic_30[name].loc[20140901:].std()
ir_30[name] = ic_30_mean[name]/ic_30_std[name]import datetime
trade_date = pd.Series(ic_30.index)
trade_date = trade_date.apply(lambda x: datetime.datetime.strptime(str(x), '%Y%m%d'))
ic_30.index = trade_date可视化比较
pd.Series(ic_30_mean).plot(kind="barh",xerr=pd.Series(ic_30_std),figsize=(15,5))<matplotlib.axes._subplots.AxesSubplot at 0x18b5240e828>
print(ic_30_mean["equal_weight"])
print(ic_30_mean["ic_weight"])
print(ic_30_mean["pe"])0.11461587810097988
0.10435470638726971
0.1067541063545408
pd.Series(ir_30).plot(kind="barh",figsize=(15,5))<matplotlib.axes._subplots.AxesSubplot at 0x18b5086f9b0>
print(ir_30["equal_weight"])
print(ir_30["ic_weight"])
print(ir_30["pe"])0.5528241142805751
0.48673093039146453
0.4986503963545165
ic_30[["equal_weight","ic_weight","pe"]].plot(kind="line",figsize=(15,5),)<matplotlib.axes._subplots.AxesSubplot at 0x18b4ff6bf28>
ic_30.loc[datetime.date(2017,1,3):,][["equal_weight","ic_weight","pe"]].plot(kind="line",figsize=(15,5),)<matplotlib.axes._subplots.AxesSubplot at 0x18b4fcc6940>
import matplotlib.pyplot as plt
from jaqs_fxdayu.research.signaldigger.analysis import analysis
from jaqs_fxdayu.research import SignalDigger
obj = SignalDigger()
obj.process_signal_before_analysis(signal=comb_factors["equal_weight"],
price=dv.get_ts("close_adj"),
high=dv.get_ts("high_adj"), # 可为空
low=dv.get_ts("low_adj"),# 可为空
n_quantiles=5,# quantile分类数
mask=mask,# 过滤条件
can_enter=can_enter,# 是否能进场
can_exit=can_exit,# 是否能出场
period=30,# 持有期
benchmark_price=dv.data_benchmark, # 基准价格 可不传入，持有期收益（return）计算为绝对收益
commission = 0.0008,
)
obj.create_full_report()
plt.show()Nan Data Count (should be zero) : 0; Percentage of effective data: 56%
Value of signals of Different Quantiles Statistics
min max mean std count count %
quantile
1 0.000000 0.538462 0.103221 0.060599 53388 20.145655
2 0.180602 0.628763 0.308244 0.060081 53003 20.000377
3 0.371237 0.695652 0.510054 0.059086 52990 19.995472
4 0.565217 0.849498 0.708706 0.057843 53003 20.000377
5 0.755853 1.000000 0.904544 0.056265 52626 19.858119
Figure saved: E:\2018_Course\HighSchool\Final\5_因子研发工具实操Richard\returns_report.pdf
Information Analysis
ic
IC Mean 0.120
IC Std. 0.205
t-stat(IC) 17.957
p-value(IC) 0.000
IC Skew -0.128
IC Kurtosis -0.719
Ann. IR 0.584
Figure saved: E:\2018_Course\HighSchool\Final\5_因子研发工具实操Richard\information_report.pdf
<matplotlib.figure.Figure at 0x18b506e5048>
print(analysis(obj.signal_data,is_event=False,period=30)){'ic': return_ic upside_ret_ic downside_ret_ic
IC Mean 1.199666e-01 -0.025340 2.590128e-01
IC Std. 2.054836e-01 0.203069 1.710594e-01
t-stat(IC) 1.795679e+01 -3.838066 4.657144e+01
p-value(IC) 2.933009e-62 0.000132 6.274188e-247
IC Skew -1.281287e-01 0.369542 -4.648359e-01
IC Kurtosis -7.191112e-01 -0.728975 -1.038014e-01
Ann. IR 5.838256e-01 -0.124786 1.514168e+00, 'ret': long_ret long_short_ret top_quantile_ret bottom_quantile_ret \
t-stat 5.009707 11.970392 27.353712 -21.213261
p-value 0.000000 0.000000 0.000000 0.000000
skewness -0.049712 0.305483 2.104621 1.352262
kurtosis 4.585943 1.671780 13.188646 6.368882
Ann. Ret 0.034979 0.085021 0.097492 -0.105558
Ann. Vol 0.075573 0.076875 0.287875 0.404814
Ann. IR 0.462853 1.105960 0.338663 -0.260757
occurance 946.000000 946.000000 52626.000000 53388.000000
tmb_ret all_sample_ret
t-stat 12.331461 -2.447519
p-value 0.000000 0.014390
skewness 0.245940 1.495618
kurtosis 1.438639 9.227692
Ann. Ret 0.203389 -0.004609
Ann. Vol 0.178518 0.341296
Ann. IR 1.139320 -0.013503
occurance 946.000000 265010.000000 , 'space': long_space top_quantile_space bottom_quantile_space \
Up_sp Mean 0.128720 0.126579 0.136456
Up_sp Std 0.085865 0.140843 0.158185
Up_sp IR 1.499101 0.898725 0.862640
Up_sp Pct5 0.041861 0.004368 0.004635
Up_sp Pct25 0.076599 0.036603 0.038489
Up_sp Pct50 0.103032 0.085234 0.090519
Up_sp Pct75 0.143446 0.165011 0.176643
Up_sp Pct95 0.331293 0.391718 0.421635
Up_sp Occur 946.000000 52626.000000 53388.000000
Down_sp Mean -0.137471 -0.108866 -0.191665
Down_sp Std 0.088789 0.202787 0.282507
Down_sp IR -1.548294 -0.536849 -0.678443
Down_sp Pct5 -0.343109 -0.384268 -1.000800
Down_sp Pct25 -0.147208 -0.097760 -0.171967
Down_sp Pct50 -0.109330 -0.046730 -0.086841
Down_sp Pct75 -0.089392 -0.019800 -0.039916
Down_sp Pct95 -0.063714 -0.003965 -0.008188
Down_sp Occur 946.000000 52626.000000 53388.000000
tmb_space all_sample_space
Up_sp Mean 0.320615 0.130071
Up_sp Std 0.162529 0.143170
Up_sp IR 1.972659 0.908508
Up_sp Pct5 0.152553 0.004635
Up_sp Pct25 0.215860 0.038288
Up_sp Pct50 0.269578 0.088612
Up_sp Pct75 0.355337 0.172445
Up_sp Pct95 0.648456 0.395784
Up_sp Occur 946.000000 265010.000000
Down_sp Mean -0.247340 -0.152250
Down_sp Std 0.110376 0.253253
Down_sp IR -2.240885 -0.601178
Down_sp Pct5 -0.477266 -1.000800
Down_sp Pct25 -0.304182 -0.133021
Down_sp Pct50 -0.211650 -0.063759
Down_sp Pct75 -0.167795 -0.027543
Down_sp Pct95 -0.121475 -0.005339
Down_sp Occur 946.000000 265010.000000 }
进一步测试下等权合成因子的绝对收益效果
obj.process_signal_before_analysis(signal=comb_factors["equal_weight"],
price=dv.get_ts("close_adj"),
high=dv.get_ts("high_adj"), # 可为空
low=dv.get_ts("low_adj"),# 可为空
n_quantiles=5,# quantile分类数
mask=mask,# 过滤条件
can_enter=can_enter,# 是否能进场
can_exit=can_exit,# 是否能出场
period=30,# 持有期
#benchmark_price=dv.data_benchmark, # 基准价格 可不传入，持有期收益（return）计算为绝对收益
commission = 0.0008,
)
obj.create_full_report()
plt.show()Nan Data Count (should be zero) : 0; Percentage of effective data: 56%
Value of signals of Different Quantiles Statistics
min max mean std count count %
quantile
1 0.000000 0.538462 0.103221 0.060599 53388 20.145655
2 0.180602 0.628763 0.308244 0.060081 53003 20.000377
3 0.371237 0.695652 0.510054 0.059086 52990 19.995472
4 0.565217 0.849498 0.708706 0.057843 53003 20.000377
5 0.755853 1.000000 0.904544 0.056265 52626 19.858119
Figure saved: E:\2018_Course\HighSchool\Final\5_因子研发工具实操Richard\returns_report.pdf
Information Analysis
ic
IC Mean 0.120
IC Std. 0.205
t-stat(IC) 17.957
p-value(IC) 0.000
IC Skew -0.128
IC Kurtosis -0.719
Ann. IR 0.584
Figure saved: E:\2018_Course\HighSchool\Final\5_因子研发工具实操Richard\information_report.pdf
<matplotlib.figure.Figure at 0x18b51124cf8>
excel_data = obj.signal_data[obj.signal_data['quantile']==5]["quantile"].unstack().replace(np.nan, 0).replace(5, 1)
print (excel_data.head())
excel_data.to_excel('./equal_weight_quantile_5.xlsx')symbol 000001.SZ 000002.SZ 000012.SZ 000024.SZ 000027.SZ 000039.SZ \
trade_date
20140103 0.0 1.0 0.0 0.0 0.0 0.0
20140106 0.0 1.0 0.0 0.0 0.0 0.0
20140107 0.0 1.0 0.0 0.0 0.0 0.0
20140108 0.0 1.0 0.0 0.0 0.0 0.0
20140109 0.0 1.0 0.0 0.0 0.0 0.0
symbol 000063.SZ 000069.SZ 000100.SZ 000157.SZ ... 601919.SH \
trade_date ...
20140103 1.0 0.0 0.0 1.0 ... 0.0
20140106 1.0 0.0 0.0 1.0 ... 0.0
20140107 0.0 0.0 0.0 1.0 ... 0.0
20140108 0.0 0.0 0.0 1.0 ... 0.0
20140109 0.0 0.0 0.0 1.0 ... 0.0
symbol 601928.SH 601933.SH 601939.SH 601988.SH 601989.SH 601991.SH \
trade_date
20140103 0.0 0.0 0.0 1.0 0.0 1.0
20140106 0.0 0.0 0.0 1.0 0.0 1.0
20140107 0.0 0.0 0.0 0.0 0.0 1.0
20140108 0.0 0.0 0.0 0.0 0.0 0.0
20140109 0.0 0.0 0.0 1.0 0.0 0.0
symbol 601992.SH 601997.SH 601998.SH
trade_date
20140103 1.0 0.0 0.0
20140106 1.0 0.0 0.0
20140107 1.0 0.0 0.0
20140108 1.0 0.0 0.0
20140109 1.0 0.0 0.0
[5 rows x 244 columns]