How do I create a series based on two pandas dataframe bool columns?

    round_up  round_down is_round_up is_round_down High     Low
0   0.75       0.7       False       True          0.70532  0.69818
1   0.75       0.7       False       True          0.70196  0.67268
2   0.75       0.7       False       True          0.71243  0.69938
3   0.75       0.7       False       True          0.70226  0.69884
4   0.75       0.7       False       True          0.70292  0.69952
5   0.75       0.7       True        True          0.75100  0.69000

Desired output is a series with round_up if is_round_up or round_down if is_round_down or if both True select is_round_up.

0 0.70
1 0.70
2 0.70
3 0.70
4 0.70
5 0.75

Test data

df = pd.DataFrame({'round_up': {0: 0.75,
  1: 0.75,
  2: 0.75,
  3: 0.75,
  4: 0.75,
  5: 0.75},
 'round_down': {0: 0.70,
  1: 0.70,
  2: 0.70,
  3: 0.70,
  4: 0.70,
  5: 0.70},
 'is_round_up': {0: False, 1: False, 2: False, 3: False, 4: False, 5:True},
 'is_round_down': {0: True, 1: True, 2: True, 3: True, 4: True, 5:True},
 'High': {0: 0.70532, 1: 0.70196, 2: 0.71243, 3: 0.70226, 4: 0.70292, 5:0.751},
 'Low': {0: 0.69818, 1: 0.67268, 2: 0.69938, 3: 0.69884, 4: 0.69952, 5:0.69}})

Edit: To clarify if both columns are False I would expect to see Nan for that row.

CodePudding user response：

Select the columns in order of choice (up is preferred to down in case of both True), mask the False values, bfill and get the first column.

Advantage: the solution is scalable to any number of columns.

(df[['round_up', 'round_down']]
 .where(df[['is_round_up', 'is_round_down']].values)
 .bfill(axis=1)
 .iloc[:,0]
 .rename('result')
)

output:

0    0.70
1    0.70
2    0.70
3    0.70
4    0.70
5    0.75
Name: result, dtype: float64

CodePudding user response：

Use np.select ([condition], [choices])

condition=[df['is_round_up'],df['is_round_down'],df['is_round_up']|df['is_round_down']]
choice =[df['round_up'],df['round_down'],df['round_up']]
df['new']=np.select(condition, choice)



  round_up  round_down  is_round_up  is_round_down     High      Low   new
0      0.75         0.7        False           True  0.70532  0.69818  0.70
1      0.75         0.7        False           True  0.70196  0.67268  0.70
2      0.75         0.7        False           True  0.71243  0.69938  0.70
3      0.75         0.7        False           True  0.70226  0.69884  0.70
4      0.75         0.7        False           True  0.70292  0.69952  0.70
5      0.75         0.7         True           True  0.75100  0.69000  0.75