import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
#加载数据
df = pd.read_csv("data.csv")
df.head()

df['level'] = df['level'].fillna(0)
df['level'] = df['level'].apply(lambda x:0 if x==0 else int(x[0]))
df.head(10)

df['hot'] = df['hot'].apply(lambda x:float("%.2f"%float(x.split(" ")[-1])))
df.head(10)

df['province'] = df['area'].apply(lambda x:x.split("·")[0] )
df['city'] = df['area'].apply(lambda x:x.split("·")[1] )
df['mini_city'] = df['area'].apply(lambda x:x.split("·")[-1] )
del df['area']
df.head()

# 统计销量最多的前10个景点
num_top = df.sort_values(by='num',axis=0,ascending=False)
num_top = num_top.reset_index(drop=True)
plt.rcParams['font.sans-serif']=['Microsfot YaHei']
plt.rcParams['axes.unicode_minus']=False
import seaborn as sns
sns.set(font="SimHei")
sns.set_context("talk")
fig = plt.figure()
sns.barplot(num_top['name'][:10],num_top['num'][:10])
plt.xticks(rotation=90)
fig.show()

#景区评级和省份的关系
df['level_sum']=1
var = df.groupby(['province','level']).level_sum.sum()
var.unstack().plot(kind="bar",figsize=(35,10),stacked=False,color=['red','blue','green','black'])

#统计人最多的前10个5A景区
top_5A = df[df['level']==5].sort_values(by="num",axis=0,ascending=False).reset_index(drop=True)
# top_5A.head()
fit = plt.figure(figsize=(35,15))
# plt.pie(top_5A["num"][:10],labels=top_5A["name"][:10],autopct='%1.2f%%')
sns.barplot(top_5A["name"][:10],top_5A["num"][:10])
plt.title("人数最多的10个5A景区")
plt.xticks(rotation=90)
plt.show()

#统计人最少的前10个5A景区
top_5A = df[df['level']==5].sort_values(by="num",axis=0,ascending=True).reset_index(drop=True)
# top_5A.head()
fit = plt.figure(figsize=(35,15))
# plt.pie(top_5A["num"][:10],labels=top_5A["name"][:10],autopct='%1.2f%%')
sns.barplot(top_5A["name"][:10],top_5A["num"][:10])
plt.title("人数最多的10个5A景区")
plt.xticks(rotation=90)
plt.show()

#景区等级分布
plt.figure(figsize=(20,1))
sns.countplot(y='level',data=df)

# 热度分布
sns.distplot(df['hot'])
plt.xticks(rotation=25)

#价格分布
sns.distplot(df['price'])
plt.xticks(rotation=25)

#销量分析
sns.distplot(df['num'])
plt.xticks(rotation=25)

df['num_cut'] = pd.cut(df['num'],10)
plt.figure()
sns.countplot(y='num_cut',data=df)

df =df[['name','level','hot','price','num','province','city','mini_city']]
one_hot_df = pd.get_dummies(df[['province','city','mini_city']]) 
one_hot_df.head()

df = df[['level','hot','price','num']]
df = pd.merge(df,one_hot_df,left_index=True,right_index=True)
df.head()

from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error,mean_absolute_error

X = df[df.columns.difference(['price'])].values
Y = df['price'].values

x_train,x_test,y_train,y_test = train_test_split(X,Y,test_size=0.3,random_state=6)
rf = RandomForestRegressor(n_estimators=20,max_depth=7)
rf.fit(x_train,y_train)
pred = rf.predict(x_test)
print(mean_squared_error(y_test,pred))
print(mean_absolute_error(y_test,pred))

result_df = pd.DataFrame(np.concatenate((y_test.reshape(-1,1),pred.reshape(-1,1)),axis=1),columns=['y_test','y_pred'])
result_df.head()