Q.Use the DataFrame created in Question 9 above to do the following:

This continues from Question 9/10's Sales DataFrame and the Sales2 DataFrame built in Q10(g).

# (a) Add 2018 as a new column by joining Sales2 alongside Sales.
# NOTE: DataFrame.append() (as taught in S2.3.5) stacks ROWS, not columns -- appending
# Sales2 that way would create 10 NaN-heavy rows instead of a clean 2018 column. Since
# every later part of this question (d, j) needs a per-person 2018 value, the join here
# is done column-wise with pd.concat(axis=1), which is the operation the rest of the
# exercise actually depends on.
Sales = pd.concat([Sales, Sales2], axis=1)
print(Sales)

print(Sales.T)                              # b) transpose
print(Sales[2017])                          # c) sales in 2017, all persons
print(Sales.loc[['Madhu', 'Ankit'], [2017, 2018]])  # d) Madhu & Ankit, 2017 and 2018
print(Sales.loc['Shruti', 2016])            # e) Shruti's 2016 sales

# f) add salesman Sumeet
Sales.loc['Sumeet'] = [196.2, 37800, 52000, 78438, 38852]

# g) delete the 2014 column
Sales = Sales.drop(columns=[2014])

# h) delete Kinshuk's row
Sales = Sales.drop(index=['Kinshuk'])

# i) rename Ankit -> Vivaan, Madhu -> Shailesh
Sales = Sales.rename(index={'Ankit': 'Vivaan', 'Madhu': 'Shailesh'})

# j) update Shailesh's 2018 sale to 100000
Sales.loc['Shailesh', 2018] = 100000
print(Sales)

# k) write to CSV with no row/column labels
Sales.to_csv('SalesFigures.csv', header=False, index=False)

# l) read it back, then restore the row/column labels
SalesRetrieved = pd.read_csv('SalesFigures.csv', header=None)
print(SalesRetrieved)
SalesRetrieved.columns = Sales.columns
SalesRetrieved.index = Sales.index
print(SalesRetrieved)

Output (key steps):

a)             2014   2015    2016   2017    2018
   Madhu      100.5  12000   20000  50000  160000
   Kusum      150.8  18000   50000  60000  110000
   Kinshuk    200.9  22000   70000  70000  500000
   Ankit    30000.0  30000  100000  80000  340000
   Shruti   40000.0  45000  125000  90000  900000

c)  Madhu      50000
    Kusum      60000
    Kinshuk    70000
    Ankit      80000
    Shruti     90000
    Name: 2017, dtype: int64

d)          2017    2018
    Madhu   50000  160000
    Ankit   80000  340000

e) 125000 …

a) Combine Sales and Sales2. Section 2.3.5 taught DataFrame.append(), but append() stacks the SECOND DataFrame's rows below the first, taking the union of columns — since Sales2 has a different column (2018) instead of matching rows, appending it that way would produce 10 rows full of NaN, not a clean fifth column. Because every later part of this question (d, j) needs each person's individual 2018 figure, Sales2 is joined here as a new COLUMN using pd.concat([Sales, Sales2], axis=1), giving a single 5-row × 5-column DataFrame with 2014–2018.

b) Transpose — Sales.T swaps rows and columns: the years become row labels and the sales-person names become column labels.

c) 2017 sales for everyone — Sales[2017] selects that one column, returned as a Series indexed by sales-person.

d) Madhu & Ankit, 2017 and 2018 — Sales.loc[['Madhu', 'Ankit'], [2017, 2018]] selects that 2×2 block by label.

e) Shruti's 2016 sales — Sales.loc['Shruti', 2016] selects a single cell: 125000.

f) Add Sumeet — Sales.loc['Sumeet'] = [...] adds a new row; assigning to a label that doesn't yet exist in the index appends it as a new row (this is how the textbook's own running example adds rows to a DataFrame).

g) Delete the 2014 column — Sales.drop(columns=[2014]).

h) Delete Kinshuk's row — Sales.drop(index=['Kinshuk']).

i) Rename labels — Sales.rename(index={'Ankit': 'Vivaan', 'Madhu': 'Shailesh'}) relabels those two rows; every other row label is unchanged.

j) Update a single value — Sales.loc['Shailesh', 2018] = 100000 overwrites just that one cell.

k) Export — Sales.to_csv('SalesFigures.csv', header=False, index=False) writes only the raw values, one row per line, comma-separated — no header row and no row-label column.

l) Re-import — since the file has no header, pd.read_csv(..., header=None) reads it back with default integer row/column labels; assigning SalesRetrieved.columns = Sales.columns and SalesRetrieved.index = Sales.index restores the original labels.

# (a) Add 2018 as a new column by joining Sales2 alongside Sales.
# NOTE: DataFrame.append() (as taught in S2.3.5) stacks ROWS, not columns -- appending
# Sales2 that way would create 10 NaN-heavy rows instead of a clean 2018 column. Since
# every later part of this question (d, j) needs a per-person 2018 value, the join here
# is done column-wise with pd.concat(axis=1), which is the operation the rest of the
# exercise actually depends on.
Sales = pd.concat([Sales, Sales2], axis=1)
print(Sales)

print(Sales.T)                              # b) transpose
print(Sales[2017])                          # c) sales in 2017, all persons
print(Sales.loc[['Madhu', 'Ankit'], [2017, 2018]])  # d) Madhu & Ankit, 2017 and 2018
print(Sales.loc['Shruti', 2016])            # e) Shruti's 2016 sales

# f) add salesman Sumeet
Sales.loc['Sumeet'] = [196.2, 37800, 52000, 78438, 38852]

# g) delete the 2014 column
Sales = Sales.drop(columns=[2014])

# h) delete Kinshuk's row
Sales = Sales.drop(index=['Kinshuk'])

# i) rename Ankit -> Vivaan, Madhu -> Shailesh
Sales = Sales.rename(index={'Ankit': 'Vivaan', 'Madhu': 'Shailesh'})

# j) update Shailesh's 2018 sale to 100000
Sales.loc['Shailesh', 2018] = 100000
print(Sales)

# k) write to CSV with no row/column labels
Sales.to_csv('SalesFigures.csv', header=False, index=False) …