我有一个要向量化的功能:
import pandas as pd
import numpy as np
import random
import statsmodels.api as sm
data = pd.DataFrame({
'state': ['a', 'b', 'c']*200,
'read': [random.uniform(10,50) for i in range(600)],
'write': [random.uniform(0,10) for i in range(600)],
'cansu': [random.uniform(11,20) for i in range(600)],
'brink': [random.uniform(2,10) for i in range(600)]
})
loop = pd.DataFrame({
'state': ['a','a','c','b','c'],
'x': [1,2,3,2,4],
'y': [2,3,4,4,1]
})
def regress(z,x,y):
X = data.query("state==@z").iloc[:,x].values
X = sm.add_constant(X)
Y = data.query("state==@z").iloc[:,y].values
result = sm.OLS(Y,X).fit()
return result.params[1]
我知道我可以使用apply, list comprehensions, itertools, map, filter, reduce, np.vectorize, etc.
和所有很酷的功能。但是,我希望能够执行以下操作:
loop['slope'] = regress(loop['state'].values, loop['x'].values, loop['y'].values)
目前不起作用。这可能吗?如果是,该如何重写或修改我的函数以使其成为可能?
答案 0 :(得分:0)
以这种方式尝试
与您的代码相同:
{
"main": "node_modules/expo/AppEntry.js",
"scripts": {
"start": "expo start",
"android": "expo start --android",
"ios": "expo start --ios",
"web": "expo start --web",
"eject": "expo eject"
},
"dependencies": {
"expo": "~37.0.3",
"react": "~16.9.0",
"react-dom": "~16.9.0",
"react-native": "https://github.com/expo/react-native/archive/sdk-37.0.1.tar.gz",
"react-native-web": "~0.11.7"
},
"devDependencies": {
"babel-preset-expo": "~8.1.0",
"@babel/core": "^7.8.6"
},
"private": true
}
在列表中执行:
import statsmodels.api as sm
data = pd.DataFrame({
'state': ['a', 'b', 'c']*200,
'read': [random.uniform(10,50) for i in range(600)],
'write': [random.uniform(0,10) for i in range(600)],
'cansu': [random.uniform(11,20) for i in range(600)],
'brink': [random.uniform(2,10) for i in range(600)]
})
loop = pd.DataFrame({
'state': ['a','a','c','b','c'],
'x': [1,2,3,2,4],
'y': [2,3,4,4,1]
})
def regress(z,x,y):
X = data.query("state==@z").iloc[:,x].values
X = sm.add_constant(X)
Y = data.query("state==@z").iloc[:,y].values
result = sm.OLS(Y,X).fit()
return result.params[1]