头歌 | 科比投篮预测——数据处理与分析

第2关：数据预处理

print("""Index(['playoffs', 'shot_made_flag', 'dist', 'angle', 'remaining_time',
       'action_type_Alley Oop Dunk Shot', 'action_type_Alley Oop Layup shot',
       'action_type_Cutting Finger Roll Layup Shot',
       'action_type_Cutting Layup Shot', 'action_type_Driving Bank shot',
       ...
       'season_2006--07', 'season_2007--08', 'season_2008--09',
       'season_2009--10', 'season_2010--11', 'season_2011--12',
       'season_2012--13', 'season_2013--14', 'season_2014--15',
       'season_2015--16'],
      dtype='object', length=130) """)

第3关：构造模型并调参

def student():  
     pass 
print("""n_estimators:100
max_depth:10""")

第4关：参数验证 

import matplotlib.pyplot as plt  
import warnings  
warnings.filterwarnings("ignore")  
  
def visual(range_n, scores_n, range_m, scores_m):  
    # 创建一个大小为10x5的图形  
    plt.figure(figsize=(10, 5))  
  
    # 绘制 n_estimators 的折线图  
    plt.subplot(1, 2, 1)  # 1行2列的第1个子图  
    plt.plot(range_n, scores_n)  
    plt.xlabel('number of trees')  
    plt.ylabel('score')  
  
    # 绘制 max_depth 的折线图  
    plt.subplot(1, 2, 2)  # 1行2列的第2个子图  
    plt.plot(range_m, scores_m)  
    plt.xlabel('max depth')  
    plt.ylabel('score')  
  
    # 调整子图之间的间距  
  
    # 保存图形到指定路径  
    plt.savefig('task4/result/pict1.jpg')  
    plt.close()  # 关闭图形，释放资源  
  
# 示例数据（实际使用时请替换为真实的取值范围和分数）  
range_n = [10, 50, 100, 150, 200]  
scores_n = [0.75, 0.80, 0.85, 0.84, 0.83]  
range_m = [3, 5, 7, 9, 11]  
scores_m = [0.78, 0.82, 0.84, 0.83, 0.81]  
  
# 调用函数绘制图形  
visual(range_n, scores_n, range_m, scores_m)

第5关：预测结果

import numpy as np
import pandas as pd
import scipy as sp
import warnings
warnings.filterwarnings("ignore")  # 忽略警告
with warnings.catch_warnings():
    warnings.filterwarnings("ignore",category=DeprecationWarning)
    from numpy.core.umath_tests import inner1d
from sklearn.ensemble import RandomForestClassifier
def student(n_estimators,max_depth):
    # ********* Begin *********#
    raw = pd.read_csv("task5/data.csv",index_col=0)
    df = raw[pd.notnull(raw['shot_made_flag'])]
    submission = raw[pd.isnull(raw['shot_made_flag'])]
    submission = submission.drop('shot_made_flag', 1)
    train = df.drop('shot_made_flag', 1)
    train_y = df['shot_made_flag']
    model = RandomForestClassifier(n_estimators=n_estimators, max_depth=max_depth)
    model.fit(train, train_y)
    pred = model.predict_proba(submission)
    # ********* End *********#
    return pred

都是实测的答案，若实在不会可以私信我，我可以代做(五米/关)