For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/
Calculating RSI mismatch with Ninjatrader/Excel
-
Hi All,
I started to compare back test results in backtrader with ninjatrader and found that the RSI was different across the two platforms. I then used excel to calculate RSI and found that the ninjatrader results were much closer. Any thoughts?
Here is how I have implemented RSI:
self.rsi = bt.indicators.RSI_SMA(self.data.close, period=self.params.period) params = ( ("period", 14), ("rsi_low", 30), ("rsi_high", 70), ) Relative Strength Index Date Close Change Gain Loss Avg Gain Avg Loss RS 14-Day RSI Backtrader 20170110 0000 10.47441656 nan 20170110 0015 10.41867814 -0.06 0.06 nan 20170110 0030 10.45 0.03 0.03 nan 20170110 0045 10.475 0.03 0.03 nan 20170110 0100 10.42873472 -0.05 0.05 nan 20170110 0115 10.48 0.05 0.05 nan 20170110 0130 10.41867859 -0.06 0.06 nan 20170110 0145 10.44251791 0.02 0.02 nan 20170110 0200 10.43046593 -0.01 0.01 nan 20170110 0215 10.42365979 -0.01 0.01 nan 20170110 0230 10.4397672 0.02 0.02 nan 20170110 0245 10.45 0.01 0.01 nan 20170110 0300 10.39934476 -0.05 0.05 nan 20170110 0315 10.48598604 0.09 0.09 nan 20170110 0330 10.4225 -0.06 0.06 0.02 0.02 0.82 45.20 45.18518519 20170110 0345 10.45 0.03 0.03 0.02 0.02 0.92 48.04 53.0214425 20170110 0400 10.451 0.00 0.00 0.02 0.02 0.93 48.15 50.10351967 20170110 0415 10.45 0.00 0.00 0.02 0.02 0.92 48.05 47.27668845 20170110 0430 10.45 0.01 0.02 0.92 48.05 52.54237288 20170110 0445 10.49999991 0.05 0.05 0.02 0.01 1.17 53.88 52.42718447 20170110 0500 10.45000018 -0.05 0.05 0.02 0.02 0.93 48.07 53.86533666 20170110 0515 10.45000005 0.00 0.00 0.01 0.02 0.93 48.07 50.92838196 20170110 0530 10.45000001 0.00 0.00 0.01 0.01 0.93 48.07 52.74725275 20170110 0545 10.45000001 0.01 0.01 0.93 48.07 53.63128492 20170110 0600 10.47363907 0.02 0.02 0.01 0.01 1.06 51.40 54.64480874 20170110 0615 10.45 -0.02 0.02 0.01 0.01 0.93 48.08 50 20170110 0630 10.38078549 -0.07 0.07 0.01 0.02 0.66 39.94 47.73869347 20170110 0645 10.37896211 0.00 0.00 0.01 0.02 0.66 39.74 32.90734824 20170110 0700 10.39299999 0.01 0.01 0.01 0.02 0.73 42.04 44.48669202 20170110 0715 10.33099753 -0.06 0.06 0.01 0.02 0.55 35.59 29.96632997 20170110 0730 10.39999996 0.07 0.07 0.01 0.02 0.84 45.60 43.01369863 20170110 0745 10.32100859 -0.08 0.08 0.01 0.02 0.62 38.27 35.44018059 20170110 0800 10.25 -0.07 0.07 0.01 0.03 0.50 33.11 30.54474708 20170110 0815 10.31106544 0.06 0.06 0.02 0.02 0.68 40.53 32 20170110 0830 10.29851782 -0.01 0.01 0.01 0.02 0.65 39.56 34.49691992 20170110 0845 10.38999754 0.09 0.09 0.02 0.02 0.97 49.13 44.80968858 20170110 0900 10.26039929 -0.13 0.13 0.02 0.03 0.65 39.57 36.5819209 20170110 0915 10.28092716 0.02 0.02 0.02 0.03 0.71 41.51 38.40877915 20170110 0930 10.41500297 0.13 0.13 0.03 0.02 1.10 52.29 46.48390942 20170110 0945 10.41446338 0.00 0.00 0.03 0.02 1.09 52.25 47.79411765 20170110 1000 10.3 -0.11 0.11 0.02 0.03 0.79 44.19 45.29616725 20170110 1015 10.29151733 -0.01 0.01 0.02 0.03 0.77 43.65 44.98269896 20170110 1030 10.26753076 -0.02 0.02 0.02 0.03 0.73 42.10 42.87343216 20170110 1045 10.26753076 0.02 0.03 0.73 42.10 46.13496933 20170110 1100 10.36798778 0.10 0.10 0.02 0.02 1.03 50.65 48.10874704 20170110 1115 10.39981794 0.03 0.03 0.03 0.02 1.13 53.02 54.9436796 20170110 1130 10.37865704 -0.02 0.02 0.02 0.02 1.05 51.26 58.61148198 20170110 1145 10.37866488 0.00 0.00 0.02 0.02 1.05 51.26 54.94186047 20170110 1200 10.48375725 0.11 0.11 0.03 0.02 1.44 59.09 61.84379001 20170110 1215 10.60447758 0.12 0.12 0.03 0.02 1.93 65.87 63.20987654 20170110 1230 10.62000044 0.02 0.02 0.03 0.02 2.00 66.64 75.86206897 20170110 1245 10.62522697 0.01 0.01 0.03 0.02 2.02 66.91 75.29411765 20170110 1300 10.54 -0.09 0.09 0.03 0.02 1.42 58.60 59.90491284 20170110 1315 10.57 0.03 0.03 0.03 0.02 1.53 60.46 61.81818182 20170110 1330 10.4873114 -0.08 0.08 0.03 0.02 1.14 53.34 64.86486486 20170110 1345 10.48082444 -0.01 0.01 0.02 0.02 1.12 52.82 65.07177033 20170110 1400 10.56144109 0.08 0.08 0.03 0.02 1.40 58.31 71.44948755 20170110 1415 10.54820276 -0.01 0.01 0.03 0.02 1.33 57.14 70.11494253 20170110 1430 10.55881489 0.01 0.01 0.03 0.02 1.37 57.87 65.73311367 20170110 1445 10.48833087 -0.07 0.07 0.02 0.02 1.06 51.56 56.81114551 20170110 1500 10.49210894 0.00 0.00 0.02 0.02 1.08 51.86 58.98251192 20170110 1515 10.5696268 0.08 0.08 0.03 0.02 1.36 57.71 63.50777935 20170110 1530 10.57971704 0.01 0.01 0.03 0.02 1.40 58.42 57.84313725 20170110 1545 10.6 0.02 0.02 0.02 0.02 1.49 59.87 49.609375 20170110 1600 10.5367546 -0.06 0.06 0.02 0.02 1.15 53.58 42.57602862 20170110 1615 10.49412539 -0.04 0.04 0.02 0.02 0.99 49.78 39.02847571 20170110 1630 10.49412539 0.02 0.02 0.99 49.78 45.5078125 20170110 1645 10.54204356 0.05 0.05 0.02 0.02 1.18 54.03 47.35849057 20170110 1700 10.59999996 0.06 0.06 0.02 0.02 1.41 58.59 61.18811881 20170110 1715 10.54297 -0.06 0.06 0.02 0.02 1.13 53.01 55.57553957 20170110 1730 10.53042645 -0.01 0.01 0.02 0.02 1.08 51.85 46.83026585 20170110 1745 10.53177183 0.00 0.00 0.02 0.02 1.08 51.97 48.32635983 20170110 1800 10.53177183 0.02 0.02 1.08 51.97 47.10920771 20170110 1815 10.53177183 0.02 0.02 1.08 51.97 55.55555556 20170110 1830 10.50000001 -0.03 0.03 0.02 0.02 0.94 48.35 50.94339623 20170110 1845 10.501 0.00 0.00 0.01 0.02 0.94 48.47 40.05763689 20170110 1900 10.4841257 -0.02 0.02 0.01 0.02 0.87 46.48 36.44067797 20170110 1915 10.69000001 0.21 0.21 0.03 0.01 1.88 65.24 58.33333333 20170110 1930 10.69207163 0.00 0.00 0.03 0.01 1.89 65.37 66.17954071 20170110 1945 10.7 0.01 0.01 0.02 0.01 1.93 65.90 73.1981982 20170110 2000 10.68555171 -0.01 0.01 0.02 0.01 1.78 63.97 70.96069869 20170110 2015 10.61000001 -0.08 0.08 0.02 0.02 1.22 54.92 56.99588477 20170110 2030 10.62635517 0.02 0.02 0.02 0.02 1.29 56.36 52.92792793 20170110 2045 10.64551937 0.02 0.02 0.02 0.01 1.38 58.05 62.65356265 20170110 2100 10.59000002 -0.06 0.06 0.02 0.02 1.07 51.80 56.66666667 20170110 2115 10.59007388 0.00 0.00 0.02 0.02 1.07 51.80 56.47321429 20170110 2130 10.68261187 0.09 0.09 0.02 0.02 1.51 60.11 63.95563771 20170110 2145 10.57531175 -0.11 0.11 0.02 0.02 0.98 49.46 53.31278891 20170110 2200 10.65 0.07 0.07 0.03 0.02 1.24 55.39 60.83815029 20170110 2215 10.64999997 0.00 0.00 0.02 0.02 1.24 55.39 60.78147612 20170110 2230 10.59781898 -0.05 0.05 0.02 0.02 1.02 50.58 57.85123967 20170110 2245 10.59501682 0.00 0.00 0.02 0.02 1.01 50.33 40.91778203 20170110 2300 10.64996841 0.05 0.05 0.02 0.02 1.23 55.07 46.35416667 20170110 2315 10.58174336 -0.07 0.07 0.02 0.02 0.95 48.84 40.72327044 20170110 2330 10.63890145 0.06 0.06 0.02 0.02 1.15 53.58 46.53902798 20170110 2345 10.55001523 -0.09 0.09 0.02 0.03 0.87 46.38 45.66473988 20170111 0000 10.55001523 0.02 0.02 0.87 46.38 44.37869822 20170111 0015 10.55000001 0.00 0.00 0.02 0.02 0.87 46.38 42.68292683 20170111 0030 10.6 0.05 0.05 0.02 0.02 1.04 51.00 50.76923077 20170111 0045 10.64384486 0.04 0.04 0.02 0.02 1.21 54.69 53.89048991 20170111 0100 10.59467128 -0.05 0.05 0.02 0.02 1.01 50.13 43.23076923 20170111 0115 10.6 0.01 0.01 0.02 0.02 1.02 50.61 52.28519196 20170111 0130 10.65 0.05 0.05 0.02 0.02 1.22 54.99 50 20170111 0145 10.60000005 -0.05 0.05 0.02 0.02 1.01 50.20 45.62937063 20170111 0200 10.60000005 0.02 0.02 1.01 50.20 50.19230769 20170111 0215 10.60000005 0.02 0.02 1.01 50.20 50.48355899 20170111 0230 10.57900005 -0.02 0.02 0.02 0.02 0.92 48.00 42.65010352 20170111 0245 10.57900005 0.02 0.02 0.92 48.00 49.63855422 20170111 0300 10.63829879 0.06 0.06 0.02 0.02 1.20 54.52 49.88009592 20170111 0315 10.58000001 -0.06 0.06 0.02 0.02 0.93 48.13 53.88601036 20170111 0330 10.58000001 0.02 0.02 0.93 48.13 53.88601036 20170111 0345 10.62 0.04 0.04 0.02 0.02 1.11 52.56 58.21596244 20170111 0400 10.62 0.02 0.01 1.11 52.56 52.65957447 20170111 0415 10.59 -0.03 0.03 0.02 0.02 0.96 48.93 42.54143646 20170111 0430 10.58000026 -0.01 0.01 0.01 0.02 0.91 47.74 47.67801858 20170111 0445 10.63999999 0.06 0.06 0.02 0.01 1.21 54.81 55.29100529 20170111 0500 10.68247875 0.04 0.04 0.02 0.01 1.44 59.03 54.32432432 20170111 0515 10.66961308 -0.01 0.01 0.02 0.01 1.34 57.28 60.54216867 20170111 0530 10.65 -0.02 0.02 0.02 0.01 1.20 54.63 57.10227273 20170111 0545 10.64 -0.01 0.01 0.02 0.01 1.14 53.28 55.52486188 20170111 0600 10.66999998 0.03 0.03 0.02 0.01 1.31 56.74 62.26415094 20170111 0615 10.66999998 0.02 0.01 1.31 56.74 62.26415094 20170111 0630 10.58623347 -0.08 0.08 0.01 0.02 0.84 45.75 43.43434343 20170111 0645 10.58579206 0.00 0.00 0.01 0.02 0.84 45.70 50.88757396 20170111 0700 10.64999864 0.06 0.06 0.02 0.01 1.16 53.67 58.70646766 20170111 0715 10.63 -0.02 0.02 0.02 0.01 1.05 51.15 51.30890052 20170111 0730 10.58579207 -0.04 0.04 0.01 0.02 0.85 46.01 46.00938967 20170111 0745 10.58579209 0.00 0.00 0.01 0.02 0.85 46.01 49.49494949 20170111 0800 10.59082391 0.01 0.01 0.01 0.01 0.88 46.72 51.40664962 20170111 0815 10.64218629 0.05 0.05 0.02 0.01 1.15 53.42 50.2617801 20170111 0830 10.5910003 -0.05 0.05 0.01 0.02 0.89 47.07 38.36317136 20170111 0845 10.59100059 0.00 0.00 0.01 0.02 0.89 47.07 39.57783641 20170111 0900 10.50065316 -0.09 0.09 0.01 0.02 0.61 37.85 33.40757238 20170111 0915 10.26335002 -0.24 0.24 0.01 0.04 0.32 24.36 22.15657312 20170111 0930 10.19 -0.07 0.07 0.01 0.04 0.28 21.78 16.66666667 20170111 0945 10.18307636 -0.01 0.01 0.01 0.04 0.27 21.55 16.50618982 20170111 1000 10.09613043 -0.09 0.09 0.01 0.04 0.23 18.83 16.43835616 20170111 1015 10.16097308 0.06 0.06 0.01 0.04 0.36 26.29 23.27044025 20170111 1030 10.09000042 -0.07 0.07 0.01 0.04 0.31 23.72 15.0872818 20170111 1045 10.22018135 0.13 0.13 0.02 0.04 0.56 36.07 27.52192982 20170111 1100 10.25708866 0.04 0.04 0.02 0.03 0.64 39.08 31.82320442 20170111 1115 9.9 -0.36 0.36 0.02 0.06 0.36 26.22 22.82091918 20170111 1130 10.10184103 0.20 0.20 0.03 0.05 0.63 38.53 33.24194654 20170111 1145 10.0860799 -0.02 0.02 0.03 0.05 0.61 38.00 30.47752809
-
@j45p41 said in Calculating RSI mismatch with Ninjatrader/Excel:
I started to compare back test results in backtrader with ninjatrader
Do you have the source code of Ninjatrader to understand how they have implemented RSI?
@j45p41 said in Calculating RSI mismatch with Ninjatrader/Excel:
I then used excel to calculate RSI and found that the ninjatrader results were much closer
What is closer? RSI is well defined.
@j45p41 said in Calculating RSI mismatch with Ninjatrader/Excel:
Here is how I have implemented RSI:
No. You show a series of values. That's not an implementation and gives no clues as to what you may or may not be doing (in any case the intended tabular output is badly formatted, killing the intended tabular presentation)
You may want to read something like this additionally:
And understand that many commercial platforms do whatever they want without telling you.
-
@backtrader thank you for your explanation and for taking the time to reply, I could have been more clear with my post. I didn’t realise that there were many ways of implementing RSI :). In that case is there a way of specifying how backtrder calculates the moving average?
I do think this is a great project and have invested a lot of time in it and benefitted a lot from the support on this forum. I am just trying to port my work on NT and got stuck here.
-
@j45p41 said in Calculating RSI mismatch with Ninjatrader/Excel:
In that case is there a way of specifying how backtrder calculates the moving average?
Docs - Indicator Reference - Look for
SmoothedMovingAverageThe calculation method is not a secret, it is in Welles Wilder book and even in Wikipedia and in the Reddit post linked above.