EPTi run the analyzer and get the normal pic but without 'Shortest Paths'.
how can i fix it?...there is nothing in this thread or on sc2mapster which helped me ~~
Edit: nervermind i got it now
[Tool] SC2 Map Analyzer - Page 8
| Forum Index > SC2 Maps & Custom Games |
|
Mereel
Germany895 Posts
i run the analyzer and get the normal pic but without 'Shortest Paths'. how can i fix it?...there is nothing in this thread or on sc2mapster which helped me ~~ Edit: nervermind i got it now | ||
|
dimfish
United States663 Posts
On September 30 2010 21:38 Mereel wrote: hi i got a question, i run the analyzer and get the normal pic but without 'Shortest Paths'. how can i fix it?...there is nothing in this thread or on sc2mapster which helped me ~~ Edit: nervermind i got it now Do you mind posting what the solution was, in case somebody else gets stuck? Was it just disabled in output.txt? | ||
|
Mereel
Germany895 Posts
On October 01 2010 01:26 dimfish wrote: Do you mind posting what the solution was, in case somebody else gets stuck? Was it just disabled in output.txt? yes it was | ||
|
k10forgotten
Brazil260 Posts
 If possible, draw the numbers of the shortest paths at last. The other bases can be rendered after them - on top of them - making the whole process useless. + Show Spoiler + ![[image loading]](http://img8.imageshack.us/img8/8790/paranoidandroidshortest.png) Clearly the hidden distance is over NINE THOUSAAAAND. | ||
|
Spartan
United States2030 Posts
| ||
|
dimfish
United States663 Posts
On October 02 2010 08:34 k10forgotten wrote: Just one more thing. Again about numbers, but a graphical view of them. If possible, draw the numbers of the shortest paths at last. The other bases can be rendered after them - on top of them - making the whole process useless. + Show Spoiler + Clearly the hidden distance is over NINE THOUSAAAAND. I have a better solution for you coming soon. Very clever to put starting locations at the points you want to measure to--I'm going to allow you to place the generic Point, which is the same dialog with the Starting Location, except that you name it something and you will be able to tell the map analyzer what you want to measure. The advantages are 1) you can place a Point anywhere, 2) you can leave a Point in the release version of your map, it is totally ignored in-game, and 3) you will be able to measure very specific routes by using Points as a waypoint. That sound like what you need? | ||
|
k10forgotten
Brazil260 Posts
On October 03 2010 01:42 dimfish wrote: I have a better solution for you coming soon. Very clever to put starting locations at the points you want to measure to--I'm going to allow you to place the generic Point, which is the same dialog with the Starting Location, except that you name it something and you will be able to tell the map analyzer what you want to measure. The advantages are 1) you can place a Point anywhere, 2) you can leave a Point in the release version of your map, it is totally ignored in-game, and 3) you will be able to measure very specific routes by using Points as a waypoint. That sound like what you need? Sounds really good, but my point was only about hidden numbers! Thank you for this, though! :D | ||
|
dimfish
United States663 Posts
On October 03 2010 03:34 k10forgotten wrote: Sounds really good, but my point was only about hidden numbers! Thank you for this, though! :D I understand your issue with the path numbers layered under the start location--but it is that way on purpose. Otherwise I have to do this: 1. draw shortest paths lines 2. draw start location boxes 3. draw shortest path numbers 4. draw start location names In the code right now #1 and #3 are tightly coupled together, and #2 and #4 are tightly coupled, meaning I can't just flip a few lines of code and give you what you want. Also, I never considered that someone might put down extra start locations to get other measurements, which is very clever by the way! I think my time is best spent making the new version that is better than the workaround you're using, though. Doesn't your trick actually spit out a billion base-to-base measurements but you are really interested in the "main->nat->third" or "main->nat->gold" or something like that? I'm proposing to let you write little strings like that in a config file and you'll get pretty output. It will be easy to write "main->watchtower" or "main->myOwnDefinedPoint" sort of stuff, too. Please be patient, I really do spend a ton of my free time working on this stuff, and I'll try to get you what you're looking for soon. | ||
|
k10forgotten
Brazil260 Posts
On October 03 2010 05:54 dimfish wrote: I understand your issue with the path numbers layered under the start location--but it is that way on purpose. Otherwise I have to do this: 1. draw shortest paths lines 2. draw start location boxes 3. draw shortest path numbers 4. draw start location names In the code right now #1 and #3 are tightly coupled together, and #2 and #4 are tightly coupled, meaning I can't just flip a few lines of code and give you what you want. Also, I never considered that someone might put down extra start locations to get other measurements, which is very clever by the way! I think my time is best spent making the new version that is better than the workaround you're using, though. Doesn't your trick actually spit out a billion base-to-base measurements but you are really interested in the "main->nat->third" or "main->nat->gold" or something like that? I'm proposing to let you write little strings like that in a config file and you'll get pretty output. It will be easy to write "main->watchtower" or "main->myOwnDefinedPoint" sort of stuff, too. Please be patient, I really do spend a ton of my free time working on this stuff, and I'll try to get you what you're looking for soon. But the problem will persist if you don't correct it. Even though I am interested on this Point feature, I think that fixing a bug like this (which prevents the user from seeing the number - its goal) has a higher priority in the making of this kind of application. | ||
|
Amadi
Finland139 Posts
![[image loading]](http://img69.imageshack.us/img69/8788/kittenlovethisthreadzh8.jpg) No, really. Best tool ever. | ||
|
Randomaccount#77123
United States5003 Posts
| ||
|
Randomaccount#77123
United States5003 Posts
| ||
|
Randomaccount#77123
United States5003 Posts
| ||
|
dimfish
United States663 Posts
Anyway, what I'm thinking is of doing another companion analysis called like "playable density." It's similar to openness but gets a very different answer for one-bit-of-cliff layouts. First let's consider what you highlighted. To calculate openness for a map cell you find the distance to the closet unpathable cell in any direction. In your example the openness is very different with no cliff and one bit o' cliff, but the results of a big battle are very much the same. However, in a small-scale zergling vs. marauder battle the openness difference really did mean, as you pointed out, that the mauraders could keep from being flanked in many places by either running up to the bitty cliff or off to the edges. Okay, how do you calculate playable density? Let's define a threshold, say 15, because that is longer than the range of any unit in the game and sort of defines the size of a "battlefield." Now to calculate the playable density at a map cell you count how many playable map cells are within 15 cells by a ground route. You'll find that by this calculation the two examples you showed (totally open, open with just a bit of cliff) will have very similar playable density values. This seems like it will be confusing when I unveil it. Maybe we should take the time to give openness and playable density really descriptive names and find more examples that show what they're good for and not good for. The problem with playable density right now is that it requires A TON of shortest path calculations that are not already being computed, so I'm looking at ways to reuse calculations. On top of all the other simultaneous changes I'm working on... GACK Anyway, wouldn't you like to see both calculations as companion analyses to help you understand a new map? | ||
|
Randomaccount#77123
United States5003 Posts
| ||
|
Superouman
France2195 Posts
"To get started with the new release, download and unzip it. In the release directory, edit to-analyze.txt and output.txt with paths for your system. The comments within the files explain the options and provide examples." It was far easier before, i just pasted the file i wanted to analyze into the program's folder and everything was perfect. Why is it so complicated now? | ||
|
Randomaccount#77123
United States5003 Posts
| ||
|
Superouman
France2195 Posts
On November 11 2010 02:42 Barrin wrote: @superouman Hmm... you can just drop the file into the folder in the newest unreleased versions. I'm sure it'll be fixed next release. This is exactly what i was doing but i have now three warnings. | ||
|
dimfish
United States663 Posts
On November 11 2010 02:44 Superouman wrote: This is exactly what i was doing but i have now three warnings. Hola, superouman, I see what you mean. My intention is for you to set up the config files once, to do something like this: C: --analyzer folder ----all the config files and the EXE --some other folder on your system where you keep maps ----map1 ----map2 ----map analyzer output folder And you set up the config files so maps can stay in their folders and you have the output files thrown into a single convenient location. BUT, I agree the default config files let you use it the way you are describing. I will change the defaults back to the old way, but the leave the new comments to allow customization. EDIT: my original proposal is incompatible between the versions, do this instead In the meantime, download unofficial release 1.4.6 and edit output.txt to enable the outputs you want. You are probably going to want shortest paths and openness and summary, those are usually the most helpful for me when I'm making maps. If it doesn't work, let me know. I'm pretty short on free time right now so I haven't been dropping in on TL much. If I don't answer you here, send a message to dimfish.mapper@gmail.com and I'm sure to see that. | ||
|
Randomaccount#77123
United States5003 Posts
| ||
| ||
