{ main="item name", sub="item name", ... }
Aside from that - this tool is amazing and I've had a lot of fun with it so far. Thanks for putting it together (and maintaining it!)
All Jobs Damage Simulator And Gear Sets |
||
All Jobs Damage Simulator and Gear Sets
Offline
Posts: 150
A small feature request (convenience tool, really), a button that will copy to clipboard the current gear set in GearSwap/json format:
Code { main="item name", sub="item name", ... } Aside from that - this tool is amazing and I've had a lot of fun with it so far. Thanks for putting it together (and maintaining it!) Offline
Posts: 308
Dazusu said: » A small feature request (convenience tool, really), a button that will copy to clipboard the current gear set in GearSwap/json format I've actually already added a basic version of this in the updated GUI since it was requested a while ago through PMs. The current version does not print augments yet. This is a problem for duplicate items you have in-game with multiple augment paths (such as ambuscade capes). However, gearswap should have no issue finding the correct augmented item to equip if you only have one copy (such as Odyssey equipment). As far as I can tell, everything directly related to the new GUI code is working properly, but I keep finding small issues with how its linked to the main code (such as a checkbox not applying under specific conditions) so I'm hesitant to push the update now. I'll probably release a separate "beta" repository for this version soon and maintain both versions of the GUI until I'm comfortable with the updated one. Any plans to add berserk to warrior? considering it's pretty much fulltime and such an impactful buff.
Offline
Posts: 308
Asura.Bynebill said: » Any plans to add berserk to warrior? considering it's pretty much fulltime and such an impactful buff. (Updated GUI beta version) My most recent replies in this thread have all been "this is in the new GUI", which isn't helpful if that new GUI isn't available. So I've decided to release the current version in a "beta" repository until I'm confident that the updated code has very few common bugs remaining. The GUI is still missing a few simple features, such as plotting damage distributions (which was actually the whole reason I started this project), but the most-used features are included and functioning (as far as I can tell, but I haven't tested every possible scenario). You will certainly run into bugs using this beta code. I've rewritten/cleaned the main code to treat the player stats as a proper class instance. It is now much easier to read, especially the new create_player.py file, which was the old set_stats.py file. All of the player's stats (after gear and buffs, but ignoring things like attack bonuses/penalties from WSs) are stored in the the "player.stats" dictionary, and the equipped gear in the "player.gearset" dictionary. The damage calculation algorithms should be unchanged, but damage output between the two versions of the code might be slightly different due to rounding/truncation differences. If you see any large differences between the two, then it's probably a typo and should be fixed. The updated code beta repository can be found in this new GitHub repository. You can go to the Actions page to download the GitHub generated executable and a zipped folder containing only the necessary files for running the code. I've made the following adjustments/additions to the code. The following list is likely not complete, but should cover the large changes. While rewriting the code, I corrected a lot of small issues. See below for an incomplete list of things I've fixed. Most of these are not yet fixed in the old version of the code. In addition to rewriting the main code, I've rewritten the GUI using tkinter, which is included by default with Python3, instead of PySimpleGUI. The new gui_wsdist.py code is currently over 3500 lines of poorly-written, but somehow mostly functional, code. I plan to clean up this GUI code later. For now this will have to do. The layout of the GUI has not changed significantly. The following changes have been made, many of which are very nice quality of life changes suggested by users that I would not have thought about myself. Note that the updated GUI does not support screen resolutions with height less than ~900 pixels. I will add a horizontally-oriented GUI later, but this is low priority. Additionally, I use the "vista" theme within tkinter, which means the GUI is intended for use with Windows Vista or higher. Linux, MacOS, and earlier versions of Windows do not have access to this theme and will result in a broken GUI. Apparently, the GUI looks slightly different (better even) on Windows11. Finally, I very frequently encounter small issues with the updated GUI and code, which is why I've been hesitant to upload the beta. But it appears to function well for all of the many situations I've tested (and fixed) at this point. You will almost certainly run into an error message or some other issue. When you do encounter a bug, feel free to post it here or PM me (or preferably post on the GitHub Issues page) so I can try to replicate/resolve it. I imagine there will be many small bugs so hopefully we don't need to spam this thread with small issues. Please provide as much detail as possible for each bug post (ws, spell, buffs enabled, button clicked, error message, etc). Feel free to keep posting feature requests here as well. shijin spiral ftp should be 1.5
Offline
Posts: 308
Ramuh.Austar said: » shijin spiral ftp should be 1.5 This is actually the second time you've told me this. I must have gotten busy and not fixed the issue the first time, or accidentally reverted the correction. Either way, both versions of the code now use FTP=1.5 for Shijin Spiral. Thanks. Thanks for the link to the updated/beta version. I see in your pic you list berserk as 25% and only mention warriors JP and Mythic enhancement. Berserk for warrior main job is 35%, 25% for sub.
Berserk Offline
Posts: 308
Asura.Bynebill said: » Thanks for the link to the updated/beta version. I see in your pic you list berserk as 25% and only mention warriors JP and Mythic enhancement. Berserk for warrior main job is 35%, 25% for sub. Berserk Thanks. This is an issue I would have not caught by myself. Berserk now provides WAR main with +35% (0.34765625) attack, while WAR sub gets +25% (0.25). The tooltip has also been updated. kuroki said: » how valuable and viable would a ws metric option for ws accuracy / magic accuracy be? for things like full break or tachi: ageha etc. I like this idea and it is easy enough to add. The outputs from this are not surprising, though. My first thought is to have the code maximize (hit_rate * magic_accuracy), so more magic accuracy is always good, but accuracy stops being valuable after reaching the hit rate cap. I've added Shield Break, Armor Break, Weapon Break, Full Break, Nightmare Scythe, Tachi: Ageha, Shell Crusher, Skullbreaker, Sniper Shot, and Dulling Arrow. These technically did not need to be added since the code is not maximizing damage, so it didn't need their WSC or FTP, but it doesn't hurt to have them for completeness. I've also added an Angon checkbox which shows up for all jobs and applies DEF-20% to the enemy. I will eventually swap this for a drop-down menu which will also include Defense Down from WSs, DNC steps, and BLU spells. Finally, I've included the enemy's finalized stats (after debuffs) in the "Show stats" button. This just makes it easy to confirm the effect of debuffs. Izanami said: » My first thought is to have the code maximize (hit_rate * magic_accuracy), so more magic accuracy is always good, but accuracy stops being valuable after reaching the hit rate cap. Offline
Posts: 150
Do you take donations? Got a PayPal link?
Offline
Posts: 308
Ramuh.Austar said: » Izanami said: » My first thought is to have the code maximize (hit_rate * magic_accuracy), so more magic accuracy is always good, but accuracy stops being valuable after reaching the hit rate cap. This is a good idea too. I may change the algorithm to do this later, but for now I'm not worried about it. Dazusu said: » Do you take donations? Got a PayPal link? A few people have asked me about donations, so I may consider it in the future in the sense of a "buy me a coffee" button hidden in the menu. But for now, I do not want to take donations since it would make me feel like I'm obligated to work on this project, even when I'd rather do other things. So, I dig the beta version with the new gui a lot - and, it also seems to be a fair bit faster than the old one. Thanks a lot.
However, this is how it looks like on my laptop: It's workable for some basic functionality, but most buffs are unselectable unfortunately :-\ Laptop screen is 1920x1080, Offline
Posts: 308
Lili said: » However, this is how it looks like on my laptop: It's workable for some basic functionality, but most buffs are unselectable unfortunately :-\ Laptop screen is 1920x1080, no DPI scaling, windows 10. Interesting considering I'm also using 1920x1080 Windows10. I'm not sure about my DPI scaling, though. At first glance, I think the issue might be that your Windows10 default font size is slightly larger than mine, and that font size is being inherited by the GUI for the text. Your list of gear seems to hold 13 options compared to the ~16 I see in my screenshot at the top of this page. Text field (and buttons with text) widths in Tkinter are defined in units of "average character width", rather than pixels (width=5 will hold 5 letters/numbers on average). So larger font size leads to larger text widgets, which potentially leads to widgets being pushed out of frames as you're seeing. I think the solution for this is for me to turn on the "automatically expand frames to fit their contents" option, which I remember being pretty annoying to properly use, so I chose not to use it at the time. This should be an easy fix, but the GUI code is a mess, so I'm not sure if it will be a quick fix. Essentially all rows/columns in each frame, and subframes in those frames and so on, need to be updated with a small configuration change to one parameter. I'll see what I can do with this soon. Hopefully it's a quick fix. Izanami said: » I'll see what I can do with this soon. Hopefully it's a quick fix. So, I'm a dumbass so I went to doublecheck DPI scaling, and I actually have my laptop set at 120% cuz I'm an old fart who doesn't like to use glasses. However, compatibility settings to disable scaling don't seem to be helping... at this point I'll gladly wait for your next version. Testing out the new version, the UI definitely feels better and more responsive.
One issue I noticed is it doesn't seem to be taking Footwork into account anymore on MNK. Both Tornado Kick and Dragon Kick damage are not affected by toggling it, and Anchorite's Gaiters +3 aren't increasing the WS Damage. This worked in the previous GUI, so hopefully it's not a difficult fix. Another issue is, on WAR, it appears to be using DA values over 100 when calculating time to WS. You can test this by putting on any set with >100 DA using the JSE+1 earring, looking at the time to WS and TP/round, then swapping it for the +2 and checking again. Aside from that, do you think you could expand the available buffs/debuffs to include Distract III, Frazzle III, Box Step, Quickstep and Stutter Step? I'd ask about Aria as well, but I'm not sure if we have good enough data on how that works yet. Offline
Posts: 308
Lili said: » However, compatibility settings to disable scaling don't seem to be helping... at this point I'll gladly wait for your next version. I'll probably work on this over the weekend. This is one of those GUI101 things that I should probably learn how to do anyway. Asura.Kaelann said: » One issue I noticed is it doesn't seem to be taking Footwork into account anymore on MNK. Both Tornado Kick and Dragon Kick damage are not affected by toggling it, and Anchorite's Gaiters +3 aren't increasing the WS Damage. This worked in the previous GUI, so hopefully it's not a difficult fix. Looks like Footwork was applying the bonus stats, but there were a few simple issues.
These have been fixed. Asura.Kaelann said: » Another issue is, on WAR, it appears to be using DA values over 100 when calculating time to WS. You can test this by putting on any set with >100 DA using the JSE+1 earring, looking at the time to WS and TP/round, then swapping it for the +2 and checking again. I simply forgot to limit some stats at 100%. I've added a cap to QA, TA, DA, Crit Rate, Daken, and Kick Attacks. Zanshin already had its rate capped at 100%. Asura.Kaelann said: » Aside from that, do you think you could expand the available buffs/debuffs to include Distract III, Frazzle III, Box Step, Quickstep and Stutter Step? I'd ask about Aria as well, but I'm not sure if we have good enough data on how that works yet. I'm having to redo all my macros and equipsets from scratch (don't ask), so I was considering targets and buffs for each of my jobs. For example, I'd only really use SCH in Sortie in a party, while THF is often used on older content with low or no ilvl with trusts.
I'm interested to know what ilvl targets and buffs players typically aim for with each job. If we could compile a list of common targets and buffs per job, this may help newer players using the program to create reliable sets. From my experience, on DRK I was typically aiming for i135 with BRD (+7 Honor/Victory/Advancing) and COR (+7 Chaos/Samurai), but I also had some sets for i140. I could not see the point in having any sets without those buffs present. Please share your typical habits using the program so far, and we can make a list together. Offline
Cerberus.Kylos said: » Honor/Victory/Advancing Only need two of those, so can add one minuet. Marcato honor + victory will cap magic haste. Or marcato minuet and two marches. Dodik said: » Only need two of those, so can add one minuet. Marcato honor + victory will cap magic haste. Or marcato minuet and two marches. Honor+Victory will cap even without Marcato. Depending whether you need extra acc, could be better to Marcato Minuet V than HMarch. i believe the program still only accounts for tp/second for the tp sets and doesn't include dps values, so if you're using it for that purpose then the only relevant stat for tp sets are going to be mob evasion. generally if you can hit an apex mob without major buffs of a specific challenge level, then you can hit NMs just fine with food and acc buffs, especially if 2h'd.
for example, 139 apex mobs have 1344 evasion. Ou has 1380 something(1388?) if i recall. of course there will still be some really high evasion ones for the same level. albumen is like 1776 compared to teles with 1588 evasion despite both being CL150 your accuracy will also adjust based on your party setup, unless you are the only DD you shouldn't focus on what buffs maximize your damage but the damage of the group as a whole. if your tp bonus offhanding friend needs an extra accuracy buff, you can drop that much from your sets. attack requirements will be much harder to recommend settings for since those will vary a lot more, particularly when you're limited on party comps, geo nerfs, or whatever else. Offline
Posts: 16
Valor Minuet V +8 is 223 ATK Honor March +4 is 232 ATK Edit: See below Katyl said: » Assuming capped +Song/Skill where relevant, it is never better to Marcato Valor Minuet V. Valor Minuet V +8 is 223 ATK Honor March +4 is 232 ATK Assuming you don't have Minuet merits or JP I guess, but I think it's fairly safe to assume most BRDs have Minuet JP and 5/5 Minuet merits, so the real numbers are: 248 (plus STR from empyrean armor) and 232. Offline
Gilgamesh.Maletaru said: » Honor+Victory will cap even without Marcato. Depending whether you need extra acc, could be better to Marcato Minuet V than HMarch. True, most people marcato honor march though. I tend to favour minuet for marcato because the 50% extra attack on it is really strong. Honor march attack with marcato: 348 Minuet5 with marcato and full equipment/merit buffs: 372 Not insignificant. Offline
Posts: 16
Huh, well I suppose there is a use case then.
I was under the impression the bonus Minuet Job Point category was only for the BRD from the wording compared to the Minne Job Point category. Increases physical defense granted by Minne. Increases physical attack while under the effects of Minuet. But after testing in game it does affect the song on others. Go figure. Thanks for the clarification. Edit: After checking in game it does say granted like Minne, but on the BG Job page it says under the effects of. No idea if this was ninja patched or why BG has different wording. Live Laugh FFXI KujahFoxfire said: » The difference is so small its almost always better to take the extra haste and accuracy from HM. Except we're talking about scenarios with HMarch+VMarch, so the haste difference is 0, unless you're slowed I guess? So ultimately it comes down to: Gilgamesh.Maletaru said: » Depending whether you need extra acc, could be better to Marcato Minuet V than HMarch. If you need the extra ~29 acc, Marcato your HMarch, if you don't, the optimal thing is to Marcato the Minuet. See also: if you're buffing RNG/COR who are shooting. It may not make a big difference either way (8 attack, as you said), but it's still a bigger number with no drawbacks. At least in the scenario being described. When you guys picture dmg over like 30 or 40k+, against how much def/mdef we are talking? Also, what buff/debuff are being considered on those simulations?
|
||
All FFXI content and images © 2002-2024 SQUARE ENIX CO., LTD. FINAL
FANTASY is a registered trademark of Square Enix Co., Ltd.
|