def main(args):
dates = F.get_date_list(args)
for i in range(len(dates) - 1):
d0, d1 = F.get_dates(i, dates)
print("Running for", d0)
gt = GetTopics(args, d0)
gt.import_data().vectorize().get_best().save_aspects().get_visual()
print("Topics for", d0, "found and saved")
def main(args):
F.set_up_folders(args)
dates = F.get_date_list(args)
for i in range(len(dates) - 1):
d0, d1 = F.get_dates(i, dates)
print("Running for", d0)
gt = GetTweets(args, d0, d1)
gt.get_tweets().fill_list().save()
print("Done", d0, "found", len(gt.all_info), "tweets")
def main(args):
dates = F.get_date_list(args)
for i in range(len(dates) - 1):
d0, d1 = F.get_dates(i, dates)
pt = ProcessTweets(args, d0)
pt.read_data() \
.get_user() \
.get_islanders() \
.add_dummy_cols() \
.format_time() \
.contains_pic() \
.apply_processing() \
.get_sentiment() \
.weighted_sentiment() \
.save()
def __init__(self, args):
self.args = args
self.df = F.import_all(
os.path.join(args.bucket, "season_" + str(args.season),
"processed/"))
self.inc_islander()
self.agg_df = pd.DataFrame()
self.islanders = get_islanders_s(args.season)
def add_cols(self):
self.agg_df[
"total_favs"] = self.agg_df["favs"] * self.agg_df["n_tweets"]
self.agg_df[
"total_retwe"] = self.agg_df["retwe"] * self.agg_df["n_tweets"]
self.agg_df = self.agg_df.merge(F.get_islander_df(self.args.season),
on="islander",
how="left")
self.agg_df["n_days"] = self.agg_df["dumped"] - self.agg_df["arrived"]
self.agg_df["n_days"] = np.where(
self.agg_df["dumped"] == 0,
60 - self.agg_df["arrived"],
self.agg_df["n_days"],
)
return self
def inc_islander(self, col="islanders"):
self.df[col] = self.df[col].apply(lambda x: F.str_to_list(x))
self.df["inc_islander"] = self.df[col].apply(lambda x: len(x))
self.df["inc_islander"] = np.where(self.df["inc_islander"] < 1, "No",
"Yes")
return self
def __init__(self, args, d0):
self.args = args
self.d0 = d0
self.df = None
self.islanders = F.get_islanders_s(args.season)