Open jencijanos opened 5 months ago
Are your prices very close to zero? There is some discussion on this issue #157:https://github.com/TA-Lib/ta-lib-python/issues/157On Jan 21, 2024, at 4:35 AM, jencijanos @.***> wrote: found a problem in the calculations RSI. df['RSI_14_close'] = RSI(df, timeperiod=14) give me date,open,high,low,close,RSI_14_close 2024-01-21 00:30:00,1.324e-05,1.331e-05,1.323e-05,1.327e-05, 2024-01-21 00:45:00,1.327e-05,1.33e-05,1.326e-05,1.327e-05, 2024-01-21 01:00:00,1.327e-05,1.33e-05,1.317e-05,1.324e-05, 2024-01-21 01:15:00,1.323e-05,1.325e-05,1.323e-05,1.325e-05, 2024-01-21 01:30:00,1.324e-05,1.325e-05,1.321e-05,1.325e-05, 2024-01-21 01:45:00,1.324e-05,1.327e-05,1.324e-05,1.327e-05, 2024-01-21 02:00:00,1.326e-05,1.327e-05,1.325e-05,1.326e-05, 2024-01-21 02:15:00,1.326e-05,1.327e-05,1.325e-05,1.327e-05, 2024-01-21 02:30:00,1.327e-05,1.328e-05,1.325e-05,1.327e-05, 2024-01-21 02:45:00,1.327e-05,1.327e-05,1.324e-05,1.326e-05, 2024-01-21 03:00:00,1.326e-05,1.327e-05,1.324e-05,1.327e-05, 2024-01-21 03:15:00,1.326e-05,1.329e-05,1.326e-05,1.328e-05, 2024-01-21 03:30:00,1.329e-05,1.329e-05,1.327e-05,1.328e-05, 2024-01-21 03:45:00,1.328e-05,1.33e-05,1.327e-05,1.33e-05, 2024-01-21 04:00:00,1.33e-05,1.33e-05,1.328e-05,1.33e-05,0.0 2024-01-21 04:15:00,1.33e-05,1.331e-05,1.33e-05,1.33e-05,0.0 2024-01-21 04:30:00,1.33e-05,1.33e-05,1.328e-05,1.328e-05,0.0 2024-01-21 04:45:00,1.329e-05,1.329e-05,1.326e-05,1.329e-05,0.0 2024-01-21 05:00:00,1.329e-05,1.329e-05,1.327e-05,1.327e-05,48.02516157768909 2024-01-21 05:15:00,1.328e-05,1.329e-05,1.327e-05,1.328e-05,51.66093531184409 2024-01-21 05:30:00,1.329e-05,1.329e-05,1.327e-05,1.329e-05,55.0473756916628 2024-01-21 05:45:00,1.328e-05,1.33e-05,1.327e-05,1.33e-05,58.20090420105429 2024-01-21 06:00:00,1.329e-05,1.33e-05,1.327e-05,1.328e-05,50.56124249753458 2024-01-21 06:15:00,1.328e-05,1.33e-05,1.327e-05,1.329e-05,53.82491650741884 2024-01-21 06:30:00,1.329e-05,1.329e-05,1.326e-05,1.326e-05,44.36323416932482 2024-01-21 06:45:00,1.326e-05,1.327e-05,1.326e-05,1.327e-05,47.66567506367998 2024-01-21 07:00:00,1.327e-05,1.328e-05,1.326e-05,1.328e-05,50.81004942611891 2024-01-21 07:15:00,1.327e-05,1.328e-05,1.325e-05,1.326e-05,44.98820062387063 2024-01-21 07:30:00,1.326e-05,1.328e-05,1.325e-05,1.328e-05,51.030740499714675 2024-01-21 07:45:00,1.327e-05,1.327e-05,1.325e-05,1.326e-05,45.63282909230429 2024-01-21 08:00:00,1.325e-05,1.326e-05,1.324e-05,1.324e-05,40.966190717700854 2024-01-21 08:15:00,1.325e-05,1.326e-05,1.324e-05,1.325e-05,44.0472689820242 2024-01-21 08:30:00,1.324e-05,1.326e-05,1.324e-05,1.325e-05,44.0472689820242 2024-01-21 08:45:00,1.326e-05,1.327e-05,1.325e-05,1.326e-05,47.24078857317984 2024-01-21 09:00:00,1.326e-05,1.327e-05,1.324e-05,1.325e-05,44.50524180500565 2024-01-21 09:15:00,1.324e-05,1.325e-05,1.324e-05,1.324e-05,41.89277478880581 2024-01-21 09:30:00,1.325e-05,1.325e-05,1.323e-05,1.323e-05,39.40195609170993 2024-01-21 09:45:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,43.048589907178076 2024-01-21 10:00:00,1.323e-05,1.324e-05,1.322e-05,1.324e-05,43.048589907178076 2024-01-21 10:15:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,43.048589907178076 2024-01-21 10:30:00,1.323e-05,1.325e-05,1.323e-05,1.324e-05,0.0 2024-01-21 10:45:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,0.0 2024-01-21 11:00:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,0.0 2024-01-21 11:15:00,1.324e-05,1.326e-05,1.323e-05,1.325e-05,0.0 2024-01-21 11:30:00,1.325e-05,1.326e-05,1.324e-05,1.326e-05,0.0 2024-01-21 11:45:00,1.326e-05,1.327e-05,1.325e-05,1.326e-05,0.0 2024-01-21 12:00:00,1.326e-05,1.327e-05,1.325e-05,1.325e-05,0.0 2024-01-21 12:15:00,1.326e-05,1.326e-05,1.324e-05,1.324e-05,0.0 2024-01-21 12:30:00,1.325e-05,1.325e-05,1.322e-05,1.323e-05,0.0 2024-01-21 12:45:00,1.324e-05,1.325e-05,1.322e-05,1.324e-05,0.0
zero in last 10 line. same input data calculated using pandas-ta give "other" result ... 2024-01-21 08:00:00,1.325e-05,1.326e-05,1.324e-05,1.324e-05,40.18228008822393 2024-01-21 08:15:00,1.325e-05,1.326e-05,1.324e-05,1.325e-05,43.557040811558295 2024-01-21 08:30:00,1.324e-05,1.326e-05,1.324e-05,1.325e-05,43.557040811558295 2024-01-21 08:45:00,1.326e-05,1.327e-05,1.325e-05,1.326e-05,47.02334768897121 2024-01-21 09:00:00,1.326e-05,1.327e-05,1.324e-05,1.325e-05,44.10630620385045 2024-01-21 09:15:00,1.324e-05,1.325e-05,1.324e-05,1.324e-05,41.34427175267467 2024-01-21 09:30:00,1.325e-05,1.325e-05,1.323e-05,1.323e-05,38.73219915753848 2024-01-21 09:45:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,42.63521193429178 2024-01-21 10:00:00,1.323e-05,1.324e-05,1.322e-05,1.324e-05,42.635211934291775 2024-01-21 10:15:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,42.635211934291775 2024-01-21 10:30:00,1.323e-05,1.325e-05,1.323e-05,1.324e-05,42.635211934291775 2024-01-21 10:45:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,42.635211934291775 2024-01-21 11:00:00,1.324e-05,1.325e-05,1.323e-05,1.324e-05,42.63521193429178 2024-01-21 11:15:00,1.324e-05,1.326e-05,1.323e-05,1.325e-05,47.820530475251424 2024-01-21 11:30:00,1.325e-05,1.326e-05,1.324e-05,1.326e-05,52.44935913524291 2024-01-21 11:45:00,1.326e-05,1.327e-05,1.325e-05,1.326e-05,52.449359135242915 2024-01-21 12:00:00,1.326e-05,1.327e-05,1.325e-05,1.325e-05,47.55662274186611 2024-01-21 12:15:00,1.326e-05,1.326e-05,1.324e-05,1.324e-05,43.215193045925396 2024-01-21 12:30:00,1.325e-05,1.325e-05,1.322e-05,1.323e-05,39.34691942233122 2024-01-21 12:45:00,1.324e-05,1.325e-05,1.322e-05,1.324e-05,44.679657715934646 2024-01-21 13:00:00,1.324e-05,1.325e-05,1.324e-05,1.325e-05,49.464611017551896
macos monterey Python 3.11.1 ta-lib 0.4.0 python modules: Cython==3.0.8 numpy==1.26.3 pandas==2.1.4 pycparser==2.21 requests==2.31.0 scipy==1.11.4 TA-Lib==0.4.28 pytest passed witout any error
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: @.***>
@mrjbq7 thanks for the quick response, and your link to issue #157. This "feature" make unusable implementation of TAlib in python for me... or is there some universal solution?
You can multiply your prices by 10000 and then calculate rsiYou can patch your talib the way the issue suggestsOr suggest another approach On Jan 21, 2024, at 8:15 AM, jencijanos @.***> wrote: @mrjbq7 thanks for the quick response, and your link to issue #157. This "feature" make unusable implementation of TAlib in python for me... or is there some universal solution?
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: @.***>
found a problem in the calculations RSI.
df['RSI_14_close'] = RSI(df, timeperiod=14)
give mezero in last 10 line.
same input data calculated using pandas-ta give "other" result
macos monterey Python 3.11.1 ta-lib 0.4.0
python modules: Cython==3.0.8 numpy==1.26.3 pandas==2.1.4 pycparser==2.21 requests==2.31.0 scipy==1.11.4 TA-Lib==0.4.28
pytest
passed witout any error