Project Statistics

[froggie]

^ Something's not working with your script and/or output, Eric.  I recognized several segments in Alabama that the logs recognize I've been on, but someone else is shown on your list as the sole clincher.

[yakra]

Interesting. What are some examples? I'll definitely need to look into this.

[yakra]

The code I used to produce the original log was an in-progress C++ version of siteupdate.
I added similar code to siteupdate.py to make the same log.
the log
the diff
I don't see the problem you described as you describe it, but I do see a very similar problem:
I recognized several segments in New Brunswick that the logs recognize Oscar has been on, but I am shown on my list as the sole clincher.

In the DIFF, I don't see anything in Alabama. I see two three possibilities:
1. Your segments are ones that you didn't explicitly .list, but acquired via a concurrency with another route.
2. A similar problem exists in the original siteupdate.py.
3. I goofed and didn't run my program with up-to-date UserData.

Whatever's going on here, thank you for the bug report. Some detective work is in order.
Maybe even a thread split...

[froggie]

Yes, the problem I saw is similar in nature to what you saw with NB and Oscar.  Regarding your possibilities, it's not #1..I did explicitly list the routes in my list (the first example I noticed being AL 10 near the Mississippi line).

[yakra]

Yes, the problem I saw is similar in nature to what you saw with NB and Oscar.  Regarding your possibilities, it's not #1..I did explicitly list the routes in my list (the first example I noticed being AL 10 near the Mississippi line).

Indeed. I see in froggie.list:
AL AL10 MS/AL US43
...and mapcat is listed as the pioneer on AL AL10 ComSt LinAve for example, even in the Python version.
This suggests to me that there's something fishy in the original siteupdate.py. The code I added was short & simple & straightforward, and I don't see any room for something to have gone wrong with this small addition...

Code: [Select]

# after processing traveler list files

segment_count = 0
print(et.et() + "writing pioneers.log",end="",flush=True)
pioneerslog = open('pioneers.log','w')
for h in highway_systems:
    for r in h.route_list:
        for seg in r.segment_list:
            segment_count += 1
    if len(seg.clinched_by) == 1:
                pioneerslog.write(str(seg) + " " + seg.clinched_by[0].traveler_name + "\n")
    print(".",end="",flush=True)
print("")
pioneerslog.write(str(segment_count) + " total segments.\n")
pioneerslog.close()

sys.exit()

# before reading updates.csv

[yakra]

D'OH!
Somehow I skipped getting my UserData repo up-to-date.
I pulled in the latest changes from GitHub, and saw 115 additions to froggie.list.
This should cure the Alabama problems (I'm going to test again), but I think some other bugs will still remain.

Edit: OK, try this:
New Python log (consider this the definitive one)
New C++ log (this one's buggy)
Their diff (I still have stuff to work on...)

Our westernmost pioneer on AL10 is now froggie, starting at AL69.

[yakra]

Here are the current numbers of lines per .list file.  Note that some will be invalid lines, comments, or blank lines, so consider them a bit on the high side in many cases.  The average number of lines per file is 1235.8.

Code: [Select]

    1027 bickendan.list
     955 yakra.list
      66 whopperman.list


Hm. Out of curiosity, how did you compile this list?

• yakra.list is correct, at 955 lines.
  
• bickendan.list has 1028 lines, with no newline @ EOF.
  
• whopperman.list is a unique, interesting case. Note how GitHub doesn't display new line numbers for most new lines. Most of the newlines use only the 0x0D character, as opposed to the Unix 0x0A or DOS/Windows 0x0D0A. I have to wonder if whopperman initially drafted this file on classic Mac OS. I'm counting 828 lines (791 good lines). GitHub shows 67. With the absence of a newline @ EOF, we're down to 66 as shown above.

[Jim]

Here are the current numbers of lines per .list file.  Note that some will be invalid lines, comments, or blank lines, so consider them a bit on the high side in many cases.  The average number of lines per file is 1235.8.

Code: [Select]

    1027 bickendan.list
     955 yakra.list
      66 whopperman.list


Hm. Out of curiosity, how did you compile this list?

• yakra.list is correct, at 955 lines.
  
• bickendan.list has 1028 lines, with no newline @ EOF.
  
• whopperman.list is a unique, interesting case. Note how GitHub doesn't display new line numbers for most new lines. Most of the newlines use only the 0x0D character, as opposed to the Unix 0x0A or DOS/Windows 0x0D0A. I have to wonder if whopperman initially drafted this file on classic Mac OS. I'm counting 828 lines (791 good lines). GitHub shows 67. With the absence of a newline @ EOF, we're down to 66 as shown above.
  

In the list_files directory,

Code: [Select]

wc -l * | sort -nr

[yakra]

wc --help
sort --help
Boom! Got it.

[Jim]

wc --help
sort --help
Boom! Got it.

If course you meant

man wc
man sort

[yakra]

Never really been a fan of man pages, for some reason. I'll just take the "digest version" of a --help, most of the time.

[yakra]

What's the largest WPT file, by filesize?
If that's not convenient to answer, which WPT file has the most waypoints?
tx.us087.wpt, at 26026 bytes, has 282 points.
Looking to see if there's anything larger, unless anyone else comes up with an answer first.

Edit:
largest WPT by filesize is /home/yakra/TravelMapping/HighwayData/hwy_data/RUS/asiah/rus.ah006irk.wpt, at 58882 bytes.
largest WPT by point count is /home/yakra/TravelMapping/HighwayData/hwy_data/RUS/asiah/rus.ah006irk.wpt, at 886 points.

The largest .list file, meanwhile, was much easier to find, with all the .lists in the same directory.
It's vespertine.list, at 195814 bytes and 849 lines.

[oscar]

The largest .list file, meanwhile, was much easier to find, with all the .lists in the same directory.
It's vespertine.list, at 195814 bytes and 849 lines.

vespertine.list's log file shows 8434 good lines, plus a few "bad lines", mainly for Mississippi state routes for which there isn't a preview or active system. Two other list files are over 7000 good lines.

As for wpt files, BC 97 and CA 1 were once competitive for "largest file", and may still be in North America. But both were recently streamlined, and neither comes close to rus.ah006irk.wpt.

[yakra]

Oops. I meant 8499 lines IIRC. Total line count including blanks and comments.

[michih]

More than 1.2 million miles (1,200,964 mi) in preview + active systems now.