def __init__(self, topics, path='pubmed.idx'):
"""
Initializes the class. Topics must be either a string or list. Path
is optional.
"""
if isinstance(topics, str):
self.topics = map(lambda x: x.strip(), topics.split(','))
elif isinstance(topics, list):
self.topics = topics
else:
raise Exception("Topics must be a string or list.")
self.searcher = Searcher(path)
self.offset = None
self.num_months = None
class Trends:
def __init__(self, topics, path='pubmed.idx'):
"""
Initializes the class. Topics must be either a string or list. Path
is optional.
"""
if isinstance(topics, str):
self.topics = map(lambda x: x.strip(), topics.split(','))
elif isinstance(topics, list):
self.topics = topics
else:
raise Exception("Topics must be a string or list.")
self.searcher = Searcher(path)
self.offset = None
self.num_months = None
def histogram(self, num_bins=25):
"""
Generates a histogram based on the hits found and when they occurred.
"""
self.num_bins = num_bins
dates = self._materialize_dates()
self.offset = self._date_to_num_months(min(self._flatten(dates)))
# Pre-populate the bins with zeroes.
bins = [[0 for j in range(self.num_bins)] for i in range(len(self.topics))]
i = 0
# Fill the bins!
for date_set in dates:
for d in date_set:
bins[i][self._date_to_bin_num(d)] += 1
i += 1
# Returns not only the bins, but also the maximum number of bins. This
# is only temporary for the Google Charting API.
return (bins, max(self._flatten(bins)))
def google_charts_api_url(self, hist):
colours = ",".join(["%06x" % randint(0, 16777215) for x in range(len(topics))])
data = "|".join(map(lambda x: ",".join(map(str, x)), hist[0]))
url = "http://chart.apis.google.com/chart?chxr=0,0,%d&chxt=y&chbh=a&chs=600x400&cht=bvg&chco=%s&chds=0,%d&chdl=%s&chd=t:%s" % (hist[1], colours, hist[1], "|".join(topics), data)
return url
def _flatten(self, lol):
"""
Flattens a list of lists. Only works on 2D lists.
"""
return [item for sublist in lol for item in sublist]
def _materialize_dates(self):
"""
Creates a list of lists of dates. The items in the first dimension
correspond to the topics. The second dimension corresponds to the
dates returned by the search of that topic.
"""
i = 0
all_dates = [[] for j in range(len(self.topics))]
for topic in self.topics:
dates = []
for result in self.searcher.search(topic):
dates.append(date(int(result.document.get_value(2)),
int(result.document.get_value(1)), 1))
all_dates[i] = dates
i += 1
return all_dates
def _date_to_bin_num(self, d):
"""
Determines which bin to place the given date into.
"""
return (self._date_to_num_months(d) - self.offset) % self.num_bins
def _date_to_num_months(self, d):
"""
Determines how many months since the beginning the given date is.
"""
return d.year * 12 + d.month
