Discuit user domination stats from data dump

discuit_domination.py

# disc domination
import pandas

fullDF = pandas.read_csv("d:/docs/download/discuitdump/concat.csv")
fullDF = fullDF.query("deleted != True")[["disc", "user"]]
discUser = (fullDF
  .groupby(["disc", "user"], as_index = False)
  .size()
  .rename(columns = {"size": "userItems"}))
disc = (discUser[["disc", "userItems"]]
  .groupby("disc")
  .sum()
  .rename(columns = {"userItems": "discItems"}))
dominators = disc.merge(discUser, on = "disc").sort_values(["disc", "userItems"], ascending = [True, False])
dominators["userPCT"] = 100 * dominators["userItems"] / dominators["discItems"]
dominators = dominators.drop_duplicates(["disc"], keep = "first")

print(dominators\
  .query("(userPCT >= 50) & (discItems >= 100)")
  .sort_values("discItems", ascending = False)
  .to_markdown(index = False))

print(dominators\
  .query("discItems >= 1000")
  .sort_values("discItems", ascending = False)
  .to_markdown(index = False))



##############################################################

# diversity of interests (# discs activity per user)

fullDF = pandas.read_csv("d:/docs/download/discuitdump/concat.csv")
fullDF = fullDF.query("deleted != True")[["disc", "user", "postType"]]
userDiscCount = (fullDF
  .drop_duplicates()
  .groupby("user", as_index = False)
  .size())
userDiscCount["rank"] = userDiscCount["size"].rank(method = "min", ascending = False)
userDiscCount = userDiscCount.sort_values("rank")
print(userDiscCount.query("rank <= 10").to_markdown(index = False))

userDiscCount_Posts = (fullDF
  .query("postType != 'c'")
  .drop_duplicates()
  .groupby("user", as_index = False)
  .size())
userDiscCount_Posts["rank"] = userDiscCount_Posts["size"].rank(method = "min", ascending = False)
userDiscCount_Posts = userDiscCount_Posts.sort_values("rank")
print(userDiscCount_Posts.query("rank <= 10").to_markdown(index = False))

userDiscs = (userDiscCount
  .rename(columns = {"size": "numDiscs"})
  .sort_values("numDiscs")
  .groupby("numDiscs", as_index = False)
  .size()
  .rename(columns = {"size": "users"}))
userDiscs["cumulativeUsers"] = userDiscs["users"].cumsum()
totalUsers = userDiscs.iloc[-1]["cumulativeUsers"]
userDiscs["percent"] = 100 * userDiscs["users"] / totalUsers
userDiscs["cumulativePercent"] = userDiscs["percent"].cumsum()
print(userDiscs.iloc[:25].to_markdown(index = False))