选择带条件的列数据并将其移动到新列

Question

我有一个如下所示的数据帧。你知道吗

T$QOOR
   3
  14
  12
  -6
 -19
   9

我想把正数和负数移到新的列中。你知道吗

sls_item['SALES'] = sls_item['T$QOOR'].apply(lambda x: x if x >= 0 else 0)
sls_item['RETURN'] = sls_item['T$QOOR'].apply(lambda x: x*-1 if x < 0 else 0)

结果如下。你知道吗

T$QOOR    SALES    RETURN    
   3        3         0
  14       14         0
  12       12         0
  -6        0        -6
 -19        0       -19
   9        9         0

除了使用apply之外，还有更好更干净的方法吗？你知道吗

Answer 1

用^{}和 ^{}，也^{}用于乘以-1：

sls_item['SALES'] = sls_item['T$QOOR'].clip_lower(0)
sls_item['RETURN'] = sls_item['T$QOOR'].clip_upper(0).mul(-1)
print (sls_item)
   T$QOOR  SALES  RETURN
0       3      3       0
1      14     14       0
2      12     12       0
3      -6      0       6
4     -19      0      19
5       9      9       0

使用^{}或^{}：

sls_item['SALES'] = sls_item['T$QOOR'].where(lambda x: x >= 0, 0)
sls_item['RETURN'] = sls_item['T$QOOR'].where(lambda x: x < 0, 0) * -1
print (sls_item)
   T$QOOR  SALES  RETURN
0       3      3       0
1      14     14       0
2      12     12       0
3      -6      0       6
4     -19      0      19
5       9      9       0

---

mask = sls_item['T$QOOR'] >=0
sls_item['SALES'] = np.where(mask, sls_item['T$QOOR'], 0)
sls_item['RETURN'] = np.where(~mask, sls_item['T$QOOR'] * -1, 0)
print (sls_item)
   T$QOOR  SALES  RETURN
0       3      3       0
1      14     14       0
2      12     12       0
3      -6      0       6
4     -19      0      19
5       9      9       0