Create list type new column based on division operation of the existing columns in pandas-CodePudding

I have a data frame

df = pd.DataFrame([["X",62,5],["Y",16,3],["Z",27,4]],columns=["id","total","days"])

id total days
X   62    5
Y   16    3
Z   27    4

Divide total column by days column and Create a new column plan which is a list in which No. of elements=Divisor, and the value of elements=Quotient, if any reminder is there increase those many values from negative indexing.

Expected Output:

df_out = pd.DataFrame([["X",62,5,[12,12,12,13,13]],["Y",16,3,[5, 5, 6]],["Z",27,4,[6, 7, 7, 7]]],columns=["id","total","days","plan"])

id total days    plan
X   62    5   [12, 12, 12, 13, 13]
Y   16    3   [5, 5, 6]
Z   27    4   [6, 7, 7, 7]

How to do it in pandas?

CodePudding user response：

You can use a custom function:

def split(t, d):
    # get floor division and remainder
    x, r = divmod(t, d)
    # assign divider or divider   1
    # depending on the number of remainders
    return [x]*(d-r) [x 1]*r

df['plan'] = [split(t, d) for t, d in zip(df['total'], df['days'])]

Output:


  id  total  days                  plan
0  X     62     5  [12, 12, 12, 13, 13]
1  Y     16     3             [5, 5, 6]
2  Z     27     4          [6, 7, 7, 7]

CodePudding user response：

Mozway already provided a better solution.Yet this could be another approach with the use of costume function well with lambda.

def create_plan(plan, days, remainder):
    return [plan]*days if remainder == 0 else [plan]*(days-remainder) [plan 1]*remainder

df = pd.DataFrame([["X",62,5],["Y",16,3],["Z",27,4]],columns=["id","total","days"])

# Create plan column    
df["plan"] = df["total"] // df["days"]

# Create column for remainder
df["remainder"] = df["total"] % df["days"]

# Apply function to create final plan
df["plan"] = df.apply(lambda x: create_plan(x["plan"], x["days"], x["remainder"]), axis=1)

# Drop remainder column
df.drop("remainder", axis=1, inplace=True)
print(df)

Output:

  id  total  days                  plan
0  X     62     5  [12, 12, 12, 13, 13]
1  Y     16     3             [5, 5, 6]
2  Z     27     4          [6, 7, 7, 7]