While bored on a train ride home the other day, I threw a quick Python script together that tracks how many times each altchain appears in the top N pages on this subforum for funsies and shiggles. This number N is passed in as a command line argument, to try and keep things modular. Additionally, the script counts the number of times that the word "scam" appears in thread titles as a very basic metric to track FUD on this board.
Source code at
https://www.dropbox.com/s/8g2ye2ft3xvj5d6/track_altchain_hype.pyI use Python 2.7.3 on my machine. You will also need the
BeautifulSoup library, but it has easy_install support so that shouldn't be an issue.
Example run from 9:30 am PST today:
$>python track_altchain_hype.py 10
Parsing url...https://bitcointalk.org/index.php?board=67
Parsing url...https://bitcointalk.org/index.php?board=67.40
Parsing url...https://bitcointalk.org/index.php?board=67.80
Parsing url...https://bitcointalk.org/index.php?board=67.120
Parsing url...https://bitcointalk.org/index.php?board=67.160
Parsing url...https://bitcointalk.org/index.php?board=67.200
Parsing url...https://bitcointalk.org/index.php?board=67.240
Parsing url...https://bitcointalk.org/index.php?board=67.280
Parsing url...https://bitcointalk.org/index.php?board=67.320
Parsing url...https://bitcointalk.org/index.php?board=67.360
POPULARITY RESULTS:
('LTC', [83, 15404, 950183])
('WDC', [40, 4052, 84124])
('DGC', [28, 2318, 32873])
('PXC', [28, 1631, 18869])
('XRP', [23, 6251, 126662])
('GLD', [21, 1704, 21208])
('DVC', [20, 3692, 82914])
('YAC', [17, 5870, 129839])
('FTC', [16, 2092, 78508])
('BQC', [11, 1912, 40346])
('IXC', [9, 1397, 16831])
('SCAM_INDEX', [9, 286, 7803])
('PPC', [8, 2219, 103103])
('BTE', [8, 1306, 41215])
('PWC', [8, 1013, 16584])
('RYC', [8, 764, 18476])
('JKC', [8, 653, 15343])
('CNC', [7, 1831, 51636])
('NVC', [7, 1757, 50161])
('BTB', [7, 995, 26373])
('NBC', [5, 1306, 18739])
('FRC', [5, 645, 13167])
('TRC', [5, 623, 13955])
('FRK', [3, 581, 12025])
('MNC', [3, 422, 8220])
('BTG', [3, 352, 6042])
('NMC', [3, 11, 304])
('doubloons', [1, 256, 5098])
('SRC', [1, 0, 10])
('SPC', [0, 0, 0])
SCAM RESULTS:
('MNC', [1, u'zacho56'])
('LTC', [1, u'ecliptic'])
('WDC', [0])
('GLD', [0])
('RYC', [0])
('DVC', [0])
('PPC', [0])
('DGC', [0])
('doubloons', [0])
('PXC', [0])
('PWC', [0])
('FRC', [0])
('SRC', [0])
('BQC', [0])
('FRK', [0])
('IXC', [0])
('TRC', [0])
('NBC', [0])
('FTC', [0])
('CNC', [0])
('XRP', [0])
('YAC', [0])
('SPC', [0])
('BTE', [0])
('BTG', [0])
('NVC', [0])
('NMC', [0])
('BTB', [0])
('JKC', [0])
$>
Output for "popularity results" is in the format of [number of threads, number of replies, number of views]
Output for "scam results" is in the format of [number of threads, users who started the threads (if applicable)]
Things I'm planning to improve on:
- Correlate the "scam_index" with the names of altchains to gather data on which coins are considered scams.- Add additional weight to the counts based on the number of pages in each thread and the thread's position on the board (more visible thread position and more pages in thread == higher popularity weight).- Keep track of usernames as well to find who the main cheerleaders for each coin are. (Mostly complete, just need to display them properly)
- Write a prettyprinter to make output more self-explanatory and friendly
- Enable sorting by number of replies and views, in addition to number of threads for "popularity" output.
- (Long Term) Write a GUI and allow the output results to be sorted in real-time.
Suggestions I like:
Suggestion - compile a top 10 list of regulars who use the word scam the most.
Again, this is purely for shits and giggles; I'm not trying to make a quick buck or bash any particular coin or anything here...just display some data that I think is interesting. Questions/comments/constructive feedback/suggestions/rude remarks welcome (but maybe not so much on the rude remarks).
