如何在熊猫中转行

Question

我这里有这张桌子，我想把“每日重症监护病房占用率”作为一个列，值在行下面。“每日医院入住率”也是如此。我该怎么做？我想要建立一个只有“每日重症监护病房入住率”、“每日医院入住率相同”的新数据框架，然后将这两个数据框架合并在一起，但这似乎并不是最好的方法，我肯定有更好的方法。

entity       iso_code       date                 indicator           value
127422  United States      USA 2020-07-15       Daily ICU occupancy   9245.0
127424  United States      USA 2020-07-15  Daily hospital occupancy  33759.0
127426  United States      USA 2020-07-16       Daily ICU occupancy   9797.0
127428  United States      USA 2020-07-16  Daily hospital occupancy  38537.0
127430  United States      USA 2020-07-17       Daily ICU occupancy  10700.0

更新:只是为了更清楚我在寻找什么。我希望我的新数据表是这样的

entity       iso_code       date                Daily ICU occupancy Daily hospital occupancy                   
127422  United States      USA 2020-07-15          9245.0             33759.0
127426  United States      USA 2020-07-16          9797.0             38537.0 
127430  United States      USA 2020-07-17          10700.0            38600.0 

Answer 1

您可以尝试使用枢轴来实现这一点。修改以下代码

df.pivot_table(index=['entity','iso_code'], columns='indicator',
                     values='value').reset_index()

注我的答案是基于格式问题时复制到剪贴板。调整以适应您的用例。

indicator   entity  iso_code    ICU occupancy   hospital occupancy
0   2020-07-15  Daily   9245.0  33759.0
1   2020-07-16  Daily   9797.0  38537.0
2   2020-07-17  Daily   10700.0 NaN

Answer 2

df = pd.DataFrame({
    'entity':[127422, 127424, 127426, 127428, 127430],
    'iso_code':['United States', 'United States', 'United States', 'United States', 'United States'], 
    'date':['USA 2020-07-15', 'USA 2020-07-15', 'USA 2020-07-16', 'USA 2020-07-16', 'USA 2020-07-17'],
    'indicator':['Daily ICU occupancy', 'Daily hospital occupancy', 'Daily ICU occupancy', 'Daily hospital occupancy', 'Daily ICU occupancy'],
    'value':[9245.0, 33759.0, 9797.0, 38537.0, 10700.0]
})

print(df)

   entity       iso_code            date                 indicator    value
0  127422  United States  USA 2020-07-15       Daily ICU occupancy   9245.0
1  127424  United States  USA 2020-07-15  Daily hospital occupancy  33759.0
2  127426  United States  USA 2020-07-16       Daily ICU occupancy   9797.0
3  127428  United States  USA 2020-07-16  Daily hospital occupancy  38537.0
4  127430  United States  USA 2020-07-17       Daily ICU occupancy  10700.0

df = df.pivot(index=['iso_code', 'date'], columns='indicator', values='value').reset_index().reset_index(drop=True)

print(df)

indicator       iso_code            date  Daily ICU occupancy  \
0          United States  USA 2020-07-15               9245.0   
1          United States  USA 2020-07-16               9797.0   
2          United States  USA 2020-07-17              10700.0   

indicator  Daily hospital occupancy  
0                           33759.0  
1                           38537.0  
2                               NaN  

​

​