Allow to customize quality of pokeball throws#2286
Allow to customize quality of pokeball throws#2286GepLik wants to merge 15 commits intoPokemonGoF:devfrom GepLik:dev
Conversation
|
@Likael please check how to put all the settings you add properly inside another key in the setup. The websocket settings are a good example. |
|
@douglascamata I need your help here, I must be missing something. I can't find the websocket settings (other than the key websocket_server).
I will be happy look at the websocket settings if you can guide me to it, and revise the settings to make it look more like it. |
|
@Likael this config: Automatically translates to a variable named This way you don't need any |
|
This is great! Can't wait. Would it be possible to randomize these parameters such that Excellent throws are made 15-20% of the time, for example? |
|
@douglascamata Excellent! I must have missed it. I am on it. |
Added missing curly bracket to tasks>MoveToFort>config
…2194) * Lowered the stepsize in Spiral navigator to more accurate 70m * Moved max_steps to task configuration and changed it to diameter * Added diameter to configuration example * Bugfix * Removed max_steps from cli * Added max_steps as legacy configuration * Made step size of follow_spiral more configureable * Changed default value for diameter
…sure it's really work.
* added keep_for_evo * accounted for non evolable pokemon * additional logging * additional logging * moved get_candies to utils * disregard 2nd stage evolution pokemon * added sample configs
* Added proxy support Added proxy support from config.json and command line. with parameters: -prxip | --proxy_ip or proxy_ip from config.json for ipv4 format as string "aaa.bbb.ccc.ddd" -prxp| --proxy_port or proxy_por from config.json as int from 1 to 65536. * Added proxy support Added two additional parameters (proxy_ip and proxy_port) to add support for proxy. proxy_ip must be as string of ipv4 format: "aaa.bbb.ccc.ddd" , proxy_port must be a int from 1 to 65536, if proxy_ip is null or set to "" and proxy port is null or set to 0 proxy will be ignored and not used. * Moved proxy function to a method. Moved proxy function to a method. * Changed the name of method Changed from set_proxy_if_exists to setup_proxy
* wip: fixing imports and tests * tests are passing * add tests, and modify calls to api accordingly * fix login bug * fix rebase * fix import error * Handle ThrottlingException * fix relogging errors
|
OK rebase went wrong. Will fix that tomorrow. |
|
Closing pull request to reopen a new clean one |
WARNING:
The code below is safe to use.
However, I tried to test further the possibilities of the field normalized_hit_position, and, coincidence or not, couldn't log again with my account any more for 1 hour immediately after sending a weird value. If you want to proceed and modify normalized_hit_position to test randomization, be careful.
Short Description:
The code allows to modify the frequency of the various throw possibilities.
For example, with the following setting :
"catch_throw_parameters": {
"excellent_rate": 4,
"great_rate": 10,
"nice_rate": 4,
"normal_rate": 2,
"spin_success_rate" : 0.6,
}
The bot would have 4 chances out of 20 (4+10+4+2) to land an Excellent throw.
In the sample config, those value add up to 1 to look like percentage, but the user is free to set the values to what speaks most to him/her.
The spin_success_rate is a percentage of landing a successful spin. In the example above, the bot has 60% chance of landing a spin that will generate the bonus exp.
More information about the normalized_hit_position :
The code below should work safely as it doesn't contain the further tests i have tried, however my paranoid nature would ask for people to test with account they don't care about for 1 day before merging it. I will do so myself tomorrow too. Since the (unban?) I have been testing the code below, and got no further problem.
The value of 0 for normalized_hit_position is safe, as it was what we were sending before I correct the API call.
The value of 1.0 is also safe, as we used it forever.
I didn't test any values between 0 et 1.
Values above 1 seems to be what could have caused the potential temporary ban.
The last case don't happen with this code, as the value is set either to 0 or 1.
Fixes: