SQX 136 2022.12.23 [SEVERE BUG]: SQX builds invalid strategies that look ok but are not ok (stockpicker engine)

Ok, I see why the difference. Entry/exit signals are not evaluating on the very last bar.

I will fix that in build 137 dev2

Attached are 2 screenshots of trading lists for strategy 0.173560. As you can see there are 3 extra trades using the stock group 2022.12.30. The extra trades are in the middle of the trading period, not near the end of the test. The number of trades should be the same for both stock groups, so the 3 extra trades are somewhat a mystery.

Do you see it and do you agree please??

Another minor bug is that the trades labelled "EndTest" should be labelled "Exit Signal" instead. The reason is that the exit date is the same as the end test date (see attached screenshot).

Regards

It looks that you have modified the template file from ebook and changed Position Score to RandomCondition.

Position score can by only RandomValue, we will limit that in AW.

Second problem is that Indicator MAVP has 2 periods, Min and Max. Builder generated greater Min period than Max which is the source of the 3 extra trades. 

Talib indicators returns 0 values for some periods.

The last one problem is related to the very last bar on end of the period.

All of these bugs will be fixed in the next 137 dev2 on next Monday. 6.2.2023

Thanks for your effort to test stockpicker engine :)

Nice, thank you. Yes you are correct, I used the ebook template file and modified it.

I'm worried that the MAVP indicator is not the only indicator that has a problem. Can you check all the indicators please?

Why can position score only be RandomValue please? I think it makes sense that the position score can be a random indicator, and not just a random value (I assume that random value is just a random number).

Kind regards

Search for talib_ prefix. But of course we will double check that.

RandomValue is any block except comparisons.

Is it possible to check every indicator in the TA library before the 137 dev2 release, or will it taker longer please?

        <Item key="talib_ADOSC" name="(ADOSC) Chaikin A/D Oscillator" display="ADOSC(@Chart@#FastPeriod#,#SlowPeriod#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_APO" name="(APO) Absolute Price Oscillator" display="APO(@Chart@#FastPeriod#,#SlowPeriod#,#MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_MACD" name="(MACD) Moving Average Convergence/Divergence" display="MACD(@Chart@#FastPeriod#,#SlowPeriod#,#SignalPeriod#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_MACDEXT" name="(MACDEXT) MACD with controllable MA type" display="MACDEXT(@Chart@#FastPeriod#,#FastMAType#,#SlowPeriod#,#SlowMAType#,#SignalPeriod#,#SignalMAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_MAVP" name="(MAVP) Moving Average with Variable Period" display="MAVP(@Chart@#MinPeriod#,#MaxPeriod#,#MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_PPO" name="(PPO) Percentage Price Oscillator" display="PPO(@Chart@#FastPeriod#,#SlowPeriod#,#MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_STOCH" name="(STOCH) Stochastic" display="STOCH(@Chart@#FastK_Period#,#SlowK_Period#,#SlowK_MAType#,#SlowD_Period#,#SlowD_MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_STOCHF" name="(STOCHF) Stochastic Fast" display="STOCHF(@Chart@#FastK_Period#,#FastD_Period#,#FastD_MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_STOCHRSI" name="(STOCHRSI) Stochastic Relative Strength Index" display="STOCHRSI(@Chart@#TimePeriod#,#FastK_Period#,#FastD_Period#,#FastD_MAType#)[#Shift#]" returnType="number" categoryType="indicator">

        <Item key="talib_ULTOSC" name="(ULTOSC) Ultimate Oscillator" display="ULTOSC(@Chart@#TimePeriod1#,#TimePeriod2#,#TimePeriod3#)[#Shift#]" returnType="number" categoryType="indicator">