Question

我有以下数据框：

nationwide_measures = pd.read_sql_query("""select state,
          measure_id,
          measure_name,
          score
from timely_and_effective_care___hospital;""", conn)

我创建了这个功能：

# Function to grab measure values
def get_stats(group):
    df = pd.DataFrame({'Minimum': group.min(), 'Maximum': group.max(), 'Average': group.mean(), 'Standard Deviation': group.std()}, index = [0])
    return df

# Function output
nationwide_measure_results = nationwide_measures1['score'].groupby(nationwide_measures1['measure_id']).apply(get_stats).unstack()

输出是一个包含以下5列的数据框：

“索引”| （'平均'，0）| （'最大'，0）| （ '最小'，0）| （'标准偏差”，0）

如何更改输出以重命名6列：

“测量ID”| “测量名称”| “平均”| “最大”| “最低”| “标准偏差“

我试过了：

df = pd.DataFrame({'Minimum': group.min(), 'Maximum': group.max(), 'Average': group.mean(), 'Standard Deviation': group.std()}, index = [0], columns=["Measure ID", "Average", "Maximum", "Minimum", "Standard Deviation"])

和

df.columns = ["Measure ID", "Average", "Maximum", "Minimum", "Standard Deviation"]

在功能内部，都不起作用。

Answer 1

我们试试这个例子。

import pandas as pd
import numpy as np

df = pd.DataFrame({'state':np.random.choice(['TX','CA','NY'],100),'measure_id':np.random.randint(1,5,100),'measure_name':np.nan,'score':np.random.randint(50,100,100)})

dict = {1:'Measure A',2:'Measure B',3:'Measure C',4:'Measure D',5:'Measure E'}

df['measure_name'] = df['measure_id'].map(dict)

输入数据：

   measure_id measure_name  score state
0           3    Measure C     82    CA
1           3    Measure C     93    CA
2           4    Measure D     69    NY
3           1    Measure A     56    NY
4           4    Measure D     66    CA

df_out=(df.groupby(['measure_id','measure_name'])['score'].agg(['mean','max','min','std'])
         .rename(columns={'mean':'Average','max':'Maximum','min':'Minimum','std':'Standard Deviation'})
         .rename_axis(['Measure ID','Measure Name'])
         .reset_index())

print(df_out)

输出：

   Measure ID Measure Name    Average  Maximum  Minimum  Standard Deviation
0           1    Measure A  74.346154       99       53           13.734460
1           2    Measure B  70.720000       97       50           16.084465
2           3    Measure C  76.130435       97       51           14.943239
3           4    Measure D  77.576923       97       56           10.756107

Answer 2

首先，这是一种向数据框添加新列的方法。

df['Measure ID'] = pd.Series(df.index.values)

示例

>>> import pandas as pd >>> df = pd.DataFrame({'Minimum': [1,1], 'Maximum': [0,0], 'Average': [0,1], 'Standard Deviation': [1,324]}, index = [0,1], columns=["Average", "Maximum", "Minimum", "Standard Deviation"]) Average Maximum Minimum Standard Deviation 0 0 0 1 1 1 1 0 1 324 >>> df['Measure ID'] = pd.Series(df.index.values) Average Maximum Minimum Standard Deviation Measure ID 0 0 0 1 1 0 1 1 0 1 324 1

很难给你完全你需要什么，因为我们没有所需的全部输入。但您可以按照相同的格式添加新列

df['NEW COLUMN NAME'] = pd.Series(NEW_COLUMN_DATA)

我假设您希望以与示例相同的方式排序列，这里是如何重新排序列。

让我们说这是你的数据帧：

Average Maximum Minimum Standard Deviation Measure ID Measure Name 0 0 0 1 1 0 Place 1 1 0 1 324 1 Holder

然后我们可以这样做：

>>> cols = df.columns.tolist() ['Average', 'Maximum', 'Minimum', 'Standard Deviation', 'Measure ID', 'Measure Name'] >>> cols = cols[-2:] + cols[:-2] ['Measure ID', 'Measure Name', 'Average', 'Maximum', 'Minimum', 'Standard Deviation'] >>> df = df[cols] Measure ID Measure Name Average Maximum Minimum Standard Deviation 0 0 Place 0 0 1 1 1 1 Holder 1 0 1 324

在功能中重命名Pandas索引

2 个答案: