Module 4: Data analysis with pandas · 12 min
Filter, group and sort
big = df[df["sales"] > 200] # filter rows
print(big)
print(df["sales"].mean()) # average
by_city = df.groupby("city")["sales"].sum() # total per city
print(by_city)
print(df.sort_values("sales", ascending=False)) # biggest first
df["sales_k"] = df["sales"] / 1000 # new column
Think of groupby as: split the table into groups, calculate something for each group, then put the answers together.
Always look at your data first. Wrong results usually come from surprises in the data, not from the code.