| 23 Jun 2022 |
 Rikj000#5379 | * Select all timeframes you'd like to download with space, then hit enter to start downloading all,
auto selection based upon what's configured in mgm-config.json is a planned ToDo:
https://github.com/Rikj000/MoniGoMani/issues/261 | 15:28:53 |
 AJ_4real#6745 | Which is better after small changes, 500 epoch, or 1000 epoch | 20:17:30 |
| AJ_4real#6745 | * Which is better after small changes, 500 epoch, or 1000 epoch? | 20:17:44 |
| Rikj000#5379 | 500 will be too little, too much chance to miss out on good results.
Check here to read about my recommendations:
https://discord.com/channels/819237123009150977/825778792214691840/917374288884039700 | 21:54:15 |
| 24 Jun 2022 |
|  frankcryptic joined the room. | 05:42:30 |
| AJ_4real#6745 | Is there a way to see what the current best epoch is while still running hyperopt? | 13:44:28 |
| Rikj000#5379 | Any "best" epochs found will be printed in green to your console during the HO | 14:53:46 |
| 25 Jun 2022 |
| AJ_4real#6745 | 
Download unknown.png | 15:23:18 |
| AJ_4real#6745 | No trades were made? | 15:23:18 |
| AJ_4real#6745 | 
Download unknown.png | 16:47:03 |
| AJ_4real#6745 | What is Objective? | 16:47:06 |
 Hippocritical | objective is your hyperopt-loss | 19:26:51 |
| Hippocritical | it s just a rating system in the end | 19:27:12 |
| Rikj000#5379 | In addition to Hippocritical's answer, it's indeed just a rating system,
in general the lower/more negative it goes, the better the results in the found epoch | 21:05:43 |
| AJ_4real#6745 | Okay. | 21:06:09 |
| Rikj000#5379 | Yes that can happen,
usually this is due to mis-configuration of your mgm-config and/or weighted signal interface.
For example if you configure the search spaces too large/broad (too many options),
then HO can have a hard time finding it's first "best" epoch to go further upon.
Further sometimes signals/timeframes don't work well together.
That's why I generally recommend to first simply HO the config & timeframe that is provided by default first,
since that will guarantee you some best result epochs.
Then change a single thing in your config, and HO again,
to see if it improved things, or lead to worse results or no trades etc.
And once you finally feel confident about your HO / backtest / config, give it a final test with dry-running before attempting live | 21:15:20 |
| Rikj000#5379 | * Yes that can happen,
usually this is due to mis-configuration of your mgm-config and/or weighted signal interface.
For example if you configure the search spaces too large/broad (too many options),
then HO can have a hard time finding it's first "best" epoch to go further upon.
Further sometimes signals/timeframes/settings simply don't work well together in practice.
That's why I generally recommend to first simply HO the config & timeframe that is provided by default first,
since that will guarantee you some best result epochs.
Then change a single thing in your config, and HO again,
to see if it improved things, or lead to worse results or no trades etc.
And once you finally feel confident about your HO / backtest / config, give it a final test with dry-running before attempting live | 21:16:06 |
| Rikj000#5379 | * Yes that can happen,
usually this is due to mis-configuration of your mgm-config and/or weighted signal interface.
For example if you configure the search spaces too large/broad (too many options),
then HO can have a hard time finding it's first "best" epoch to go further upon.
Further sometimes signals/timeframes/settings simply don't work well together in practice.
That's why I generally recommend to first simply HO the default config & timeframe that is provided by MGM,
since that will guarantee you some "best" result epochs.
Then change a single thing in your config, and HO again,
to see if it improved things, or lead to worse results or no trades etc.
And once you finally feel confident about your HO / backtest / config, give it a final test with dry-running before attempting live | 21:16:49 |
| Rikj000#5379 | * Yes that can happen,
usually this is due to mis-configuration of your mgm-config and/or weighted signal interface.
For example if you configure the search spaces too large/broad (too many options),
then HO can have a hard time finding it's first "best" epoch to go further upon.
Further sometimes signals/timeframes/settings simply don't work well together in practice.
That's why I generally recommend to first simply HO the default config & timeframe that is provided by MGM,
since that will guarantee you some "best" result epochs.
Then change a single thing in your config, and HO again,
to see if it improved things, or lead to worse results or no trades etc.
Repeat that till you finally feel confident about your HO / backtest / config,
give it a final test with dry-running before attempting live | 21:17:12 |
| Rikj000#5379 | * Yes that can happen,
usually this is due to mis-configuration of your mgm-config and/or weighted signal interface.
For example if you configure the search spaces too large/broad (too many options),
then HO can have a hard time finding it's first "best" epoch to go further upon.
Further sometimes signals/timeframes/settings simply don't work well together in practice.
That's why I generally recommend to first simply HO the default config & timeframe that is provided by MGM,
since that will guarantee you some "best" result epochs.
Then change a single thing in your config, and HO again,
to see if it improved things, or lead to worse results or no trades etc.
Repeat that till you finally feel confident about your HO / backtest / config,
and then finally give it a test with dry-running before attempting live | 21:18:19 |
| Rikj000#5379 | The default mgm-config.json isn't bad by the way.. 😄
It's the best config which the MGM community and myself have found together so far,
we've tweaked it together by learning out of our hundreds of HOs | 21:21:06 |
| Rikj000#5379 | * The default mgm-config.json isn't a bad starting point by the way.. 😄
It's the best config which the MGM community and myself have found together so far,
we've tweaked it together by learning out of our hundreds of HOs | 21:21:52 |
| 29 Jun 2022 |
| frankcryptic | Redacted or Malformed Event | 01:38:36 |
|  Rikj000 banned frankcryptic (Scammer). | 04:33:45 |
| AJ_4real#6745 | What setting forces the bot to sell if the current profit for a trade drops below X percent? | 15:48:39 |
| AJ_4real#6745 | I think its stoploss_spaces.trailing_stop_positive_min_value but Idk for sure. | 15:50:49 |
| Rikj000#5379 | Yush, it should be a trailing stop positive | 16:18:49 |
| Rikj000#5379 | I recommend to also read the official freqtrade docs on stoplosses 😉 | 16:19:10 |
| AJ_4real#6745 | 
Download unknown.png | 22:06:59 |
| AJ_4real#6745 | My current worst trade is -15%, If I wanted to force sell when a trade goes below -5%, what should stoploss_spaces.trailing_stop_positive_min_value be set too? right now it's 0.05. | 22:07:00 |
