Hi @aguettinger,
There is a switch in the run command that does what you want - limits the age of the data. Try: "trade run -h" and look at the --age switch usage.
In-Development TradeDangerous: power-user trade optimizer
- Thread starter kfsone
- Start date
Hi @aguettinger,
There is a switch in the run command that does what you want - limits the age of the data. Try: "trade run -h" and look at the --age switch usage.
Great news. I have an observation about the updates. The time is being reported as UTC on your page that describes when changes were made. But I'm in the UK (currently = UTC) and your update timestamps are in the future
Great news. I have an observation about the updates. The time is being reported as UTC on your page that describes when changes were made. But I'm in the UK (currently = UTC) and your update timestamps are in the future
Argh - darn Daylight Savings...
tks - I'll sort that out.
Thanks for the replies, guys.
I always use the --age switch, but I doesn't really eliminate the problem that the searches take very long or not complete at all. While calculating the 2 or 3 hop the progress bar (enabled with -P) basically just stops. I guess the problem is that trade.py still has to look through all the data entries and make a check for the date. And with an exponential search this is just getting too much.
I will look into the --prune-hops switch, until now I always used the default.
Maybe one way to improve the search would be an optional switch while importing data with the plugin from maddavo that all entries older than a given age are stripped?
This would possibly lead to a smaller, lighter database that is much faster to process for trade.py.
I'm wondering what your PC spec is and/or whether it/Python needs tuning.
I have a DOS batch job that generates 2 hop, 3 hop and 4 hop (1 jump) trade routes for 13,37.5,42.5,76.5,105.7ly jumps where the source and destination station are the same. It does this twice. Once for stations that accept large ships, and once for stations that accept medium and small ships only. All data more than 30 days old is ignored, as are trades with narcotics.
So that's 30 runs of trade.py . It takes about 1 hour 40 minutes to run.
Last edited:
I don't know? It's a decent system. You tell me.
I'm currently in OMANI/Lovell Vision searching for some goods to trade. I tried with hops=2 and jumps=3, but didn't get a result in reasonable time. (I stopped after 10 mins).
Code:
trade.py run -vvv -P --from "OMANI/Lovell Vision" --hops 2 --jumps-per 3 --prune-hop 2 --prune-score 90 --capacity 320 --ly-per 18.00 --empty-ly 28.00 --limit 0 --credits 10000000 --age 50
* Hop 1: .........1 origins
NOTE: Pruned 318 origins
* Hop 2: ........36 origins .. 297'920-990'080cr gain, 931-3'094cr/ton
[ ] ^CTraceback (most recent call last):So I tried hops=2 jumps=2:
Code:
time trade.py run -vvv -P --from "OMANI/Lovell Vision" --hops 2 --jumps-per 2 --prune-hop 2 --prune-score 90 --capacity 320 --ly-per 18.00 --empty-ly 28.00 --credits 10000000 --age 50
* Hop 1: .........1 origins
NOTE: Pruned 117 origins
* Hop 2: ........13 origins .. 333'120-491'200cr gain, 1'041-1'535cr/ton
OMANI/Lovell Vision -> DAHAN/Nelson Works (score: 992716.799040)
Start CR: 10'000'000
Hops : 2
Jumps : 4
Gain CR : 992'418
Gain/Hop: 496'209
Final CR: 10'992'418
Load from OMANI/Lovell Vision (720ls, BMk:N, Pad:L, Plt:N, Shp:Y, Out:Y, Ref:Y):
320 x Marine Equipment 3'538cr vs 4'694cr, 1 hr vs 12 days, total: 1'132'160cr
Unload at SUDZ/Houssay Ring (41ls, BMk:N, Pad:L, Plt:N, Shp:Y, Out:Y, Ref:Y) => Gain 369'920cr (1'156cr/ton) => 10'369'920cr
Load from SUDZ/Houssay Ring (41ls, BMk:N, Pad:L, Plt:N, Shp:Y, Out:Y, Ref:Y):
140 x Fruit And Vegetables 241cr vs 2'100cr, 12 days vs 28 hrs, total: 33'740cr
106 x Grain 123cr vs 2'038cr, 12 days vs 28 hrs, total: 13'038cr
74 x Fish 346cr vs 2'498cr, 12 days vs 28 hrs, total: 25'604cr
Unload at DAHAN/Nelson Works (466ls, BMk:N, Pad:L, Plt:Y, Shp:N, Out:Y, Ref:Y) => Gain 622'498cr (1'945.31cr/ton) => 10'992'418cr
----------------------------------------------------------------------------
Finish at DAHAN/Nelson Works (466ls, BMk:N, Pad:L, Plt:Y, Shp:N, Out:Y, Ref:Y) gaining 992'418cr (1'550cr/ton) => est 10'992'418cr total
trade.py run -vvv -P --from "OMANI/Lovell Vision" --hops 2 --jumps-per 2 2 12.14s user 0.16s system 99% cpu 12.295 totalThis takes 12 seconds to complete on my system.
However it tells me I can sell Fruits and Vegetables for 2100 cr at Nelson Works in Dahan.
I'm pretty sure that is wrong.
Why does tradedangerous even consider that result? I'm pretty sure it would reject my input if I would enter this price for Fruit and Vegetables manually because the price is way off. (I think there are some boundaries for each commodity.)
However when calculating a result it ignores any boundaries.
So basically if I increase the search parameters to jumps or hops >2 the search takes forever.
Cannot speculate on the time it's taking, sorry. Have you tried without -P option?
As for the fruit and veg price, I always check on ROSS if there is something that looks suspicious to see when the update was, and what provided the update.
In your particular case, EDMC has been reporting for the last 4 days that Fruit & Veg are 2100Cr. https://ross.eddb.io/eddn/log/1597179
As for the fruit and veg price, I always check on ROSS if there is something that looks suspicious to see when the update was, and what provided the update.
In your particular case, EDMC has been reporting for the last 4 days that Fruit & Veg are 2100Cr. https://ross.eddb.io/eddn/log/1597179
Thanks!
Hmm, I've never seen "swag" used like that before. I'm more familiar with the free promo stuff at conferences, like stress balls with a company logo.
The google entry also has a series of interesting meaning. Thanks for making me discover them.
Last edited:
Some problems and temporary fixes
Hi,
I was just wondering is this tool still being supported. I've just started playing "ED - Horizons" after a year break, I was using trade-dangerous script before. But now I'm having all sorts of problems with it. Some I could fix myself some not. I'm not sure where I should post this info, may be as bug reports on github would be the right way?
The biggest gripe - it now takes ages to plan 5 hop journey towards some destination, even with pruning enabled. Not sure this can be fixed. Good job the calculation finishes at all.
Anyway, some minor bumps:
Problem: While importing from maddavos I had loads of warnings/errors about missing item names like "Water" or "Low Temperature Diamonds".
Fix: Download new "Items.csv" and put it in data folder (replacing old one).
Problem: While importing current data from edapi ("trade.py import -P edapi") sometimes script bugs out with "Invalid Key" error for ship named "Independant_Trader", may be some others too.
Fix: Add lines:
to ship_names and eddn_ship_names arrays (arrays are defined around lines 36, and 68) in ".\plugins\edapi_plug.py" file. These are just two ships missing that I've noticed, there may be others.
Please make sure you are adding "TAB" symbol before added lines (so all array entries start at same position), python is very strict about identing lines correctly.
Problem: While importing curent data from edapi - script complains about unknown items like "ANIMALMEAT" or "LOW TEMPERATURE DIAMOND".
Fix: There is a file named AltItemNames.csv, in data folder, but - I don't think it does anything right now (it would be nice if it did). For now you have to edit corrections.py file and add those lines to items array like this (around line 25):
Problem: While trying to export data to eddn I get this: "HTTPError: 426 Client Error: Upgrade Required for url: http://eddn-gateway.elite-markets.net:8080/upload/".
Fix: I'm not very good at this, but right now I'm using edmc, which does uploading automatically.
I'm not sure how internals work in tradedangerous, I might have messed up something else while trying to fix the script. Hope this will help someone, or these bugs will be fixed in future releases.
Thanks
Hi,
I was just wondering is this tool still being supported. I've just started playing "ED - Horizons" after a year break, I was using trade-dangerous script before. But now I'm having all sorts of problems with it. Some I could fix myself some not. I'm not sure where I should post this info, may be as bug reports on github would be the right way?
The biggest gripe - it now takes ages to plan 5 hop journey towards some destination, even with pruning enabled. Not sure this can be fixed. Good job the calculation finishes at all.
Anyway, some minor bumps:
Problem: While importing from maddavos I had loads of warnings/errors about missing item names like "Water" or "Low Temperature Diamonds".
Fix: Download new "Items.csv" and put it in data folder (replacing old one).
Problem: While importing current data from edapi ("trade.py import -P edapi") sometimes script bugs out with "Invalid Key" error for ship named "Independant_Trader", may be some others too.
Fix: Add lines:
Code:
'Independant_Trader': 'Keelback',
'Viper_MkIV': 'ViperMkIV',Please make sure you are adding "TAB" symbol before added lines (so all array entries start at same position), python is very strict about identing lines correctly.
Problem: While importing curent data from edapi - script complains about unknown items like "ANIMALMEAT" or "LOW TEMPERATURE DIAMOND".
Fix: There is a file named AltItemNames.csv, in data folder, but - I don't think it does anything right now (it would be nice if it did). For now you have to edit corrections.py file and add those lines to items array like this (around line 25):
Code:
items = {
"LOW TEMPERATURE DIAMOND": "Low Temperature Diamonds",
"ANIMALMEAT": "Animal Meat",
}Problem: While trying to export data to eddn I get this: "HTTPError: 426 Client Error: Upgrade Required for url: http://eddn-gateway.elite-markets.net:8080/upload/".
Fix: I'm not very good at this, but right now I'm using edmc, which does uploading automatically.
I'm not sure how internals work in tradedangerous, I might have messed up something else while trying to fix the script. Hope this will help someone, or these bugs will be fixed in future releases.
Thanks
kfsone has not updated TradeDangerous for some time. The project has been forked and is being updated by bgol.
Most if not all of the issues you have identified would be fixed by downloading the version at https://bitbucket.org/bgol/tradedangerous
Most if not all of the issues you have identified would be fixed by downloading the version at https://bitbucket.org/bgol/tradedangerous
Oh, Thanks. Just had to checkout bgol's "horizon" branch from git, so far so good. Now ... into the grinder again.
Hello,
I'm getting an error
I'm not sure what's does the error mean? Do I need to import a new file? I'm using bgol branch. It's been a while since I updated the file.
Edit: nvm finally fixed after replacing several files directly from maddavo. Don't know which file but seems to solved it.
I'm getting an error
I'm not sure what's does the error mean? Do I need to import a new file? I'm using bgol branch. It's been a while since I updated the file.
Edit: nvm finally fixed after replacing several files directly from maddavo. Don't know which file but seems to solved it.
Last edited:
There is something seriously wrong with EDDB recently.
I don't know if it is a tool that uploads data incorrectly or people are just sabotaging on purpose, but it's not just a few wrong entries.
e.g. have a look at this:
I have visited several of the systems that state they pay for Basic Medicines >1000 and none of them offered more than 400 in reality, so I guess these entries are all wrong on eddb.
I don't know if it is a tool that uploads data incorrectly or people are just sabotaging on purpose, but it's not just a few wrong entries.
e.g. have a look at this:
I have visited several of the systems that state they pay for Basic Medicines >1000 and none of them offered more than 400 in reality, so I guess these entries are all wrong on eddb.
Last edited:
That would be the System.csv . When there is an "Unrecognized SYSTEM" then it means that the prices file has a system in it that your local database does not. To resolve that you import the latest System.csv from MMS with this:
trade import -P maddavo -O Systems
BUT the next error you will probably get will be "Unrecognised STATION". So then you would need to get the Station.csv . So whenever you sense that your local data is stale, then it is usually best to use this command:
trade import -P maddavo -O Corrections,Systems,Stations,exportcsv
In general, it is best to do updates always getting the Station.csv because new ones are added all the time (mainly planetary bases). I use this command as my standard import:
trade import -P maddavo -O Corrections,Stations,exportcsv
and I add it to a batch file called madimport.bat - so I only have to type "madimport" and it executes the command.
Cheers,
Maddavo
long querys (too long actually)
dont know if its normal, or maybe i have something wrong, but
my querys lasting...
a typical query of mine would be:
trade.py run --fr anmat/rober --cr 20000000 --cap 400 --ly 18.36 --av imperialslaves,slaves -vvv --ls-m 3000 --hops 4
--jumps 2 --progress --summary -p L
maybe the query is to complex meanwhile or maybe python is not this fast
(considering the constantly rising amount of data)
someone ideas ?
thanks and happy new year
ps my pc is decent fast and im running the bgol-tradedangerous
dont know if its normal, or maybe i have something wrong, but
my querys lasting...
a typical query of mine would be:
trade.py run --fr anmat/rober --cr 20000000 --cap 400 --ly 18.36 --av imperialslaves,slaves -vvv --ls-m 3000 --hops 4
--jumps 2 --progress --summary -p L
maybe the query is to complex meanwhile or maybe python is not this fast
(considering the constantly rising amount of data)
someone ideas ?
thanks and happy new year
ps my pc is decent fast and im running the bgol-tradedangerous
Last edited:
4 hops with 2 jumps will just find to much stations to be calculated in a reasonable timeframe. You may add "--age 1" (only one day old prices allowed) to reduce the number (maybe "--no-planet" if you don't want to land on planets).
Other than that: reduce the hops and start another calculation with the end station from the first calculation. This will not give you the "best" route but it should be good enough.
Use someone still TradeDangerous and has problems with the latest version of bgol (https://bitbucket.org/bgol/tradedangerous/downloads/),
can not update the database ?
If yes, then I may have a tip. Download the latest files from maddavos site (http://www.davek.com.au/td/), item.csv etc. and exchange them
with those of bgols. This has helped me update the database without errors.
Greetings nepomuk
can not update the database ?
If yes, then I may have a tip. Download the latest files from maddavos site (http://www.davek.com.au/td/), item.csv etc. and exchange them
with those of bgols. This has helped me update the database without errors.
Greetings nepomuk
You can download the latest versions from maddavo using tradedangerous itself, no need to manually download.