Exercise - Data manipulation part 1 - Add qualifying player information

  • 3 minutes

You've identified the groups of players by examining the bimodal histograms. So now make a column to indicate whether a row represents a human or Tune Squad player. Then give each row a unique "name."

First, create the new column for the DataFrame. Create it by making a list of values for the column and then assigning the column a name.

# Initialize the list to house the player data.
pop_list = []

# If the ID number is 30 or less, it's a human player; otherwise, it's a Tune Squad player.
for id in player_df['ID']:
    if id <= 30:
        pop_list.append('player'+str(id))
    else:
        pop_list.append('tune_squad'+str(id%30))

pop_list

['player1',
 'player2',
 'player3',
 'player4',
 'player5',
 'player6',
 'player7',
 'player8',
 'player10',
 'player11',
 'player12',
 'player13',
 'player14',
 'player15',
 'player16',
 'player17',
 'player18',
 'player19',
 'player21',
 'player22',
 'player23',
 'player24',
 'player26',
 'player27',
 'player28',
 'player29',
 'tune_squad1',
 'tune_squad2',
 'tune_squad3',
 'tune_squad4',
 'tune_squad5',
 'tune_squad6',
 'tune_squad7',
 'tune_squad8',
 'tune_squad9',
 'tune_squad11',
 'tune_squad12',
 'tune_squad13',
 'tune_squad14',
 'tune_squad15',
 'tune_squad16']

Note

You took a shortcut when you used str(id%30) to create the numbers for the Tune Squad players. The modulus operator (%) will divide and return the difference.

So for every ID after 30, the code enters the else branch. For example, for 31, the else branch creates a string that's tune_squad + str(id%30), which is str(31%30). If you do the math, 31 divided by 30 is 0 remainder 1. So the player string for the row that contains ID 31 is tune_squad1.

Now add this list of strings to the DataFrame:

# Assign this list as the values for the new player column in the DataFrame.
player_df['player'] = pop_list
player_df.head()

Output

ID points possessions team_pace GP MPG TS% AST TO USG ORR DRR REBR PER player
0 1 1893.0 1251.8 97.8 63.0 33.9 0.569 17.2 11.5 26.1 4.7 23.3 7.8 10.90 player1
1 2 1386.0 1282.5 110.5 58.0 32.5 0.511 24.8 9.7 26.9 6.1 0.9 10.7 27.30 player2
2 3 1405.0 1252.3 105.8 55.0 36.3 0.605 25.7 13.9 28.1 4.5 4.9 1.8 NaN player3
3 4 1282.0 1235.9 100.7 54.0 37.6 0.636 29.5 11.0 22.3 4.8 4.6 5.6 22.35 player4
4 5 1721.0 1254.0 105.7 59.0 30.5 0.589 22.8 9.9 24.6 1.2 8.4 12.1 28.38 player5

Placing the new column last will cause problems later because the last column is the default location for new columns. So move the new column to the right of ID.

Because you have too many columns to easily type, create a list of all but the last column (now player). Use the insert() list method to move the column to the second position in the list. (Remember that Python uses zero-based indexing!)

Now use the altered list to specify the order of columns in the player_df DataFrame.

# Create list of all DataFrame column names but the last one.
column_list = list(player_df.iloc[:, :-1])

# Make player the second item in the list.
column_list.insert(1, 'player')

# Reassign the columns in the player_df DataFrame in this new order.
player_df = player_df[column_list]

# Verify that the columns are ordered the way you expect.
player_df.head()

Output

ID player points possessions team_pace GP MPG TS% AST TO USG ORR DRR REBR PER
0 1 player1 1893.0 1251.8 97.8 63.0 33.9 0.569 17.2 11.5 26.1 4.7 23.3 7.8 10.90
1 2 player2 1386.0 1282.5 110.5 58.0 32.5 0.511 24.8 9.7 26.9 6.1 0.9 10.7 27.30
2 3 player3 1405.0 1252.3 105.8 55.0 36.3 0.605 25.7 13.9 28.1 4.5 4.9 1.8 NaN
3 4 player4 1282.0 1235.9 100.7 54.0 37.6 0.636 29.5 11.0 22.3 4.8 4.6 5.6 22.35
4 5 player5 1721.0 1254.0 105.7 59.0 30.5 0.589 22.8 9.9 24.6 1.2 8.4 12.1 28.38

Now the DataFrame contains all of the player data, with a new player column that indicates whether the row of stats is for a human player or a Tune Squad player. But as you can see, you still need to impute some missing values.

© 2021 Warner Bros. Ent. All Rights Reserved.