Nhigh : do you guys want new functions ?

[eru_bahagon]

@ska : no idea what is wrong with that latest.hs... Really don't understand why this config file doesn't work the same now.

On a more joyful note :
I added Total Score Rankings and Player's Total Score to nhigh. I'll add all .sol operations too.

I want to ask you something :

Do you mind downloading your online scores each time you open nhigh if you wanna work with them (like .sol) ? (20 seconds to download, I'd say)

The problem is :
I can store the online scores for a player. I have 2 possibilities, both have disadvantages :

-either I store it in the same file as the Highscores (i.e. latest.hs)
In this case, you'll iether need to download everything together (longer) or having like 2 time-stamps for the file. Kind of weird.

-or I store it in a separate file :
In this case, you'll have two files, one with Highscores, one with your (or other players) online scores.


Concerning storing the data of all players in the Total Score Rankings (like 30 players):
I think that it is useless to have all the scores for 30 players stored, better do the Total Rankings (level and episode com together) by downloading the scores each time you want it (takes 2 minutes, progress bar and cancel button already added). A copy is stored until you close Nhigh.

[EddyMataGallos]

Thanks a lot for the new funtion, as for your question, I do prefer this:

I store it in the same file as the Highscores (i.e. latest.hs)
In this case, you'll iether need to download everything together (longer) or having like 2 time-stamps for the file. Kind of weird.
Because I have lots of diferent .hsfiles saved, and it'd a mess to have double the files.

Also, I think this:

Better do the Total Rankings (level and episode com together) by downloading the scores each time you want it (takes 2 minutes, progress bar and cancel button already added). A copy is stored until you close Nhigh.
Is better, too. But not only 30 people, since it wouldnt be accurate (you'd skip people), i think you should do it for the fisrt 100 players in the Top20 Rankings (or maybe let people choose it, dunno).

[eru_bahagon]

Is better, too. But not only 30 people, since it wouldnt be accurate (you'd skip people), i think you should do it for the fisrt 100 players in the Top20 Rankings (or maybe let people choose it, dunno).

I let people choose, like you can choose the number of player you want to take from the top-20 rankings to form the total score rankings.
100 is really long to download. 30 is enough. With 20, you have the top 15 correct. (you skip blogblob). With 30, you have the correct top 20 (but only for level), so it's fine for TRO^^
I'll release a version w/o storage by the end of the week. This version will have (using online scores):
-player's total score
-total score ranking (both ep and lvl)
-player's most improvable eps (the eps where you -or player- have the biggest diff between your level-based best and your ep score)
-beatable ep 0th (ep 0th you -or player- can take w/o improving your actual best on every lvl)
-player's non-top-20 score with difference to actual 19th
-anything I've missed ?


I'll see later about a good storage option, even if it kind of random/useless. Because actually, you have like a dictionary with players as keys and their corresponding table of scores as values. When you ask for a player's total score (either directly, or by including him into the rankings), it will download the scores only if the player is not in the dictionary. Thus, you are downloading exactly what you need.
Example : you open Nhigh and do a total score rankings with first 20 people : it download the 20.
Then you look for 999_Spring total level score (or other function). It downloads 999_Spring scores.
Then you wanna do a total score ranking with 30 people : it will download the 9 remaining.

[EddyMataGallos]

Awesome! :D Those "to be realeased" functions you quoted there are really cool, this will end up really great ;)
As for the storage system, what you said seems optimal and perfect, good luck implementing that ;D

[TheRealOne]

I'll release a version w/o storage by the end of the week. This version will have (using online scores):
-player's total score
-total score ranking (both ep and lvl)
-player's most improvable eps (the eps where you -or player- have the biggest diff between your level-based best and your ep score)
-beatable ep 0th (ep 0th you -or player- can take w/o improving your actual best on every lvl)
-player's non-top-20 score with difference to actual 19th
-anything I've missed ?

So stoked. You are the man Eru. I am not too worried about the storage of the personal .sol stuff and the total score rankings. The total score rankings will be posted on the highscore rankings thread so they will be immortalized there and maybe they can make their way in to the highscore history excel file on occasion. As for the personal stuff I don't think it is that important to have saved. I would just make it so it was saved in another file. What would be ideal is if you could save the latest.hs file like normal and then pick and choose when you save the other stuff. I don't know if that is something hard to program I have a feeling it is, but maybe not I really don't know.

Either way. you are awesome!

[eru_bahagon]

Ok, so here is a new version. I hope everything works.

Please tell if you encounter any trouble.

@TRO : thanks. :)

PS : I know the progress bar is weird, because I cannot divide the progress more than url-pages loaded. Hence, since all scores for a player are on the same page, the progress bar goes from 0% to 100% in one step. I cannot really give more precision. (What is half a page ?^^)
So for a list of 20 players (20 pages), it goes by step of 5%.

PPS: everything seems to work for me on windows (the ignore list too, funny^^)

[ska]

is this right?
Total Score for Ben_Schultz_11:
Total Level Score: 70138.250 (19:28:58.250)
Total Episode Score: 33497.100 (9:18:17.100)
Difference between both : 641.150 (0:10:41.150)

[TheRealOne]

is this right?
Total Score for Ben_Schultz_11:
Total Level Score: 70138.250 (19:28:58.250)
Total Episode Score: 33497.100 (9:18:17.100)
Difference between both : 641.150 (0:10:41.150)

Independently confirmed on two different sources other than Eru's newest nHigh. That sure is correct. What seems wrong?


Also awesome Eru I am so stoked you got this out so fast. Everything looks good right now. I'll give it a thorough test tomorrow when I have more time.

[ska]

is this right?
Total Score for Ben_Schultz_11:
Total Level Score: 70138.250 (19:28:58.250)
Total Episode Score: 33497.100 (9:18:17.100)
Difference between both : 641.150 (0:10:41.150)

Independently confirmed on two different sources other than Eru's newest nHigh. That sure is correct. What seems wrong?


Also awesome Eru I am so stoked you got this out so fast. Everything looks good right now. I'll give it a thorough test tomorrow when I have more time.

The part where I have 70k!? lol, when did that happen and how did I not notice? lol

Also, can I get a updated chronological listing of players who have reached 70k? Thanks in advance.

[EddyMataGallos]

Amazing eru, I just tested all new versions, they're great :D
I found a single bug, Player's most improvable Episodes are not wokring for me, it downlaods the scores but then doesn't print anything onscreen except for "List of most improvable episode scores for EddyMataGallos:" and then just blank space.

[TheRealOne]

So I am still having issues with the ignore list. When I try and add people the ignore list from inside the the GUI it crashes after I add like 3 people. If I try amdn manually type then in from inside the GUI this fails too. When I edit the nhighlib.py file to put in them in it doesn't save their names and then when I run nHigh again it don't get any of them in my ignore list.

Anyone else having this kid of issue? Any ideas how to stop it.

[eru_bahagon]

So I am still having issues with the ignore list. When I try and add people the ignore list from inside the the GUI it crashes after I add like 3 people. If I try amdn manually type then in from inside the GUI this fails too. When I edit the nhighlib.py file to put in them in it doesn't save their names and then when I run nHigh again it don't get any of them in my ignore list.

Anyone else having this kid of issue? Any ideas how to stop it.

Apparently it happens on windows, but not on xp (everything works on xp, everything!). On mac, I can use the Tools->add option. And I don't have any Vista/Seven to run tests. I'll try to modify the GUI in order to correct this ignore list glitch. There are a lot of glitched coming from the interface. Tkinter (it's not directly python for the gui) doesn't seem to be the best multi platform gui provider.

@EMG : You don't have any score for ep 51, that's the problem. Can you submit it ? Or is it a negative one ?
The way it works suppose that when you have a level score, you also have a ep score (you submit both of them at the same time, you can't have a score on 51-1 w/o having an ep score). And you don't have any ep score for 51.
I'll add some error management anyway, to prevent such error to ruin the whole function. I'll restrict the function to find where you have an ep score. (it only uses it to compute to difference...)

[EddyMataGallos]

Oh crap, so that Ep51 error is still there. I redid it a couple of times a month ago and submitted it a lot of times, and apparently metanet servers dont want to accept that score :S I'll try to do a Top20 score there to see if it works, if not I'll contact M&R.

[eru_bahagon]

Oh crap, so that Ep51 error is still there. I redid it a couple of times a month ago and submitted it a lot of times, and apparently metanet servers dont want to accept that score :S I'll try to do a Top20 score there to see if it works, if not I'll contact M&R.

Yeah, just improve your score on ep 51. It'll take you 10 minutes copying the 19th score on each level :)
Because I probably won't be able to correct that soon.

PS : other cause of error :
If you have a ep score, but no score for one of the levels. Like for a shitty ep 83 score with a negative score on 83-3, it will crash.

[TheRealOne]

I got the ignore list to work, but had to do it a couple times, but it seems to be holding just fine now.

But I think there is more of a error with the "Player's most improvable epsiodes" because that functions doesn't work for anyone. I tried eddy, loudog004, and golfkid, none work. I know for a fact that golf has all of his episodes uploaded with no missing scores.

EDIT: Hmm... and now it works for golf. That was strange. maybe I missed spelled his name or something.

Could "player's worst level scores" and "player's possible episode 0ths" be included in the "online scores" as well so that they could be calculated with non-top-20 scores as well?

[eru_bahagon]

There is no error in my function, but it is not able to handle with metanet glitches. Eddy has not ep score for 51, and loudog004 has an ep 27 score but no level scores... I wasn't aware of such glitches when I wrote the code. The only thing I didn't take into accout is the possibility of completing an episode without having a score for every level, like having a negative score. I bet loudog004 doesn't have a negative score on 27-1,2,3,4.
I tried the function with many players like me, xaelar, lat, lim.

Also trib ep 44 is missing too.
Vankusss has no 32-0 but a 32-1,2,3, and no ep 32.
If you launch python trough a console (like it does with a .py extension), it'll give the error. KeyError 445 is ep 44 level 5, i.e. the episode score.
KeyError 320 is ep 32 lvl 0. So, you can see if you have partly submitted episode, and maybe resubmit them will solve the problem.

About your two requests. Player's possible ep 0th is already taking into account online scores, not necessarily top 20. (otherwise, almost no one could use it, because players don't have all the top 20 of an ep)
Worst level score is the difference between your top 20 score and actual 0th. For me, it seems better to first aim for top 20, and then for 0th.

What I'll probably do is :
On the Reports->player's statistics menu : put function that doesn't need internet (i.e online scores)
Put their equivalent in the online scores tab (which actually needs internet). So 'ill add there a worst ep/lvl equivalent
Decide which one to put when you right click on a player.

EDIT : I've corrected the programm such that if an ep isn't complete, it forgets about it.
I've added a "Player [player] doesn't exist", where [player] is the stupid name you tried to get a Total score or whatever for.
I've modified the display of the time for a better layout : now it displays a time over 7 characters xxx.xxx. I hope not to see level score above 1000^^
I've modified the display of a highscore entry too :


It is better ? It allows like largest/smallest spread and more to look more like columns. Players are displays over 14 characters (max is ben, emg, darkshadow).

[TheRealOne]

Yes. Awesome Eru. I had this same problem with missing scores when I made a difference function in excel. I had to develop a macro in place spaces for missing scores. I wish I had SAS on my computer because it works a lot better for things like this.

[eru_bahagon]

Here is a correct version of NHigh, such that everyone can use Possible ep 0th/Most improvable ep
(it doesn't have worst lvl/ep score for non-top-20 scores)

[EddyMataGallos]

Thanks for the corected version eru :D Its perfect now, i took ep51 before noticing this though :P
You're doing an awesome job with NHigh, i cant even think on any more stats to implement, nice job ;)

[eru_bahagon]

I've noticed the top-20 function when you right click a level doesn't work. I will correct that. (I guess it's not important and no one use it since I haven't heard of anything concerning it)

[Innuendonewb]

Is there a nice up to date .exe version available somewhere for the less technical under us?

[eru_bahagon]

There is no .exe version actually available.
Technically, the file nhigh.pyw works as a .exe (from the user perspective, you simply double click to launch). The only thing that differs is that you previously need to download and install python 2.X. If you have any trouble, feel free to ask.
Then you can create a shortcut to nhigh.pyw, change the icon with nhighico, and place it on your desktop or wherever you want.


I might release a .exe version when everything will be corrected/added.

[Innuendonewb]

dled python, it's working splendidly.

[DJ_M]

I suggest an option for us to bundle names together. Like vankusss with van_come_to_irc and such. Would be helpful if it's possible. Maybe just for 0th scores as top20s might have duplicates.

[eru_bahagon]

I suggest an option for us to bundle names together. Like vankusss with van_come_to_irc and such. Would be helpful if it's possible. Maybe just for 0th scores as top20s might have duplicates.

I actually tried to, but unsuccessfully. My main goal to do so was to provide the same rankings as TRO do.
It is really complicating the structure of the program, since each time you want to access a player, you would need to check if he has an alias.
Considering the number of functions using player names, it is really a big work for a small gain. That's why I gave up.

I added complete worst level scores list as TRO requested.
I modified the program such that it loads by default the file latest.hs of your working directory (nhigh for windows user, manually set as Documents for mac). Works fine on windows, might not on mac -but doesn't crash the program.