JAN_mean = data_select.mean('time')

首页JAN_mean = data_select.mean('time')

时间: 2024-05-18 19:12:07 浏览: 4

这是一个计算在[时间](https://geek.csdn.net/educolumn/3fb26f40673739b3a46094a10b56bc3a?spm=1055.2569.3001.10083)维度上取平均值的操作，其中`data_select`是一个xarray数据集或数据数组，`mean`是xarray库中的一个[[函数](https://geek.csdn.net/educolumn/2319d2a62911adc34b96ea572d8225a2?spm=1055.2569.3001.10083)](https://geek.csdn.net/educolumn/ba94496e6cfa8630df5d047358ad9719?dp_token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpZCI6NDQ0MDg2MiwiZXhwIjoxNzA3MzcxOTM4LCJpYXQiOjE3MDY3NjcxMzgsInVzZXJuYW1lIjoid2VpeGluXzY4NjQ1NjQ1In0.RrTYEnMNYPC7AQdoij4SBb0kKEgHoyvF-bZOG2eGQvc&spm=1055.2569.3001.10083)，用于计算平均值，`'time'`是需要取平均值的维度。具体来说，这个操作会对`data_select`在[时间](https://geek.csdn.net/educolumn/3fb26f40673739b3a46094a10b56bc3a?spm=1055.2569.3001.10083)维度上的每个值求平均，得到一个新的数据数组或数据集，其中[时间](https://geek.csdn.net/educolumn/3fb26f40673739b3a46094a10b56bc3a?spm=1055.2569.3001.10083)维度被消除了。`JAN_mean`则是这个新的数据数组或数据集的名称。

from file_define import FileReader, TextFileReader, JsonFileReaderfrom data_define import Recordtext_file_reader = TextFileReader("D:/2011年1月销售数据.txt")json_file_reader = JsonFileReader("D:/2011年2月销售数据JSON.txt")jan_data:list[Record] = text_file_reader.read_data()feb_data:list[Record] = json_file_reader.read_data()# 将2个月份的数据合并为1个list来存储all_data:list[Record] = jan_data + feb_data# 开始进行数据计算# {“2011-01-01”：1534，“2011-01-02”：300}data_dict = {}for record in all_data: if record.date in data_dict.keys(): # 当前日期已经有记录了，所以和老记录做累加即可 data_dict[record.data] += record.money else: data_dict[record.data] = record.moneyprint(data_dict)

根据你提供的代码，你创建了一个`TextFileReader`对象和一个`JsonFileReader`对象，分别用于读取文本文件和 JSON 文件。然后，你调用了它们的`read_data()`方法，并将返回的结果存储在`jan_data`和`feb_data`列表中。接下来，你将两个月份的数据合并到一个名为`all_data`的列表中。然后，你创建了一个空字典`data_dict`，用于存储日期和对应销售金额的总和。接下来，你遍历`all_data`列表中的每个记录，并将日期作为字典的键，销售金额作为对应键的值。如果日期已经存在于字典中，则将销售金额累加到对应键的值上。否则，将日期添加到字典中，并将对应的销售金额作为初始值。最后，你打印了`data_dict`字典，其中包含每个日期和对应的销售金额总和。如果你运行这段代码并提供正确的文件路径，它应该能够读取文件并计算出每个日期的销售金额总和。如果你有任何问题或需要进一步的帮助，请随时告诉我。

from file_define import from data_define import text_file_reader = TextFileReader("D:\\BaiduNetdiskDownload\\资料\\第13章资料\\2011年1月销售数据.txt")json_file_reader = JsonFileReader("D:\\BaiduNetdiskDownload\\资料\\第13章资料\\2011年2月销售数据JSON.txt")"""参数:jan_data: 处理后的一月份数据feb_data: 处理后的二月份数据"""jan_data: list[Record] = text_file_reader.read_data()feb_data: list[Record] = json_file_reader.read_data()# 两个月的数据存放在：all_data 中 list类型all_data: list[Record] = jan_data + feb_datadata_dict: dict[str, int] = {}for record in all_data: if record.date not in data_dict.keys(): data_dict[record.date] = record.money print(f"data_dict:{data_dict}\trecord.money{record.money}") else: data_dict[record.date] = data_dict[record.date] + record.money 如何理解for循环的代码

这段代码使用了一个 `for` 循环来遍历 `all_data` 列表中的每个元素 `record`。在循环中，它首先检查 `record.d

相关推荐

openjdk-8u41-b04-linux-x64-14_jan_2020.tar.gz

openjdk：openjdk-8u41-b04-linux-x64-14_jan_2020.tar.gz

jan24.mdl.zip_jan 24_made

made complete direct power control of doubly fed induction generator.

CanRAP_Data_Collection_Analysis_Jan_2009.pdf

请优化以下代码：metss_interp = ERA5['metss_interp']ERA5_mu_Jan1 = metss_interp[:,:,0:120:12].mean(2).T#mean(2)的意思是把第三个维度的数据进行平均，原来是(144,96,120),取平均之后就是(144,96)ERA5_mu_July = metss_interp[:,:,6:120:12].mean(2).Tmntss_interp = ERA5['mntss_interp']ERA5_mv_Jan = mntss_interp[:,:,0:120:12].mean(2).TERA5_mv_July = mntss_interp[:,:,6:120:12].mean(2).Tmslhf_interp = ERA5['mslhf_interp']msshf_interp = ERA5['msshf_interp']# ERA5_mo = np.sqrt(pow(metss_interp, 2)+pow(mntss_interp,2))# ERA5_mo_Jan = ERA5_mo[:,:,0:120:12].mean(2).T# ERA5_mo_July = ERA5_mo[:,:,6:120:12].mean(2).TERA5_SH_Jan = msshf_interp[:,:,0:120:12].mean(2)ERA5_SH_Jan = -ERA5_SH_Jan.TERA5_SH_July = msshf_interp[:,:,6:120:12].mean(2)ERA5_SH_July = -ERA5_SH_July.TERA5_LH_Jan = mslhf_interp[:,:,0:120:12].mean(2)ERA5_LH_Jan = -ERA5_LH_Jan.TERA5_LH_July = mslhf_interp[:,:,6:120:12].mean(2)ERA5_LH_July = -ERA5_LH_July.T

'ERA5_mu_Jan1': (metss_interp[:,:,0:120:12].mean(2).T), 'ERA5_mu_July': (metss_interp[:,:,6:120:12].mean(2).T), 'ERA5_mv_Jan': (mntss_interp[:,:,0:120:12].mean(2).T), 'ERA5_mv_July': (mntss_interp...

import matplotlib.pyplot as pltfrom mpl_toolkits. basemap import Basemapimport xarray as xrimport pandas as pdimport numpy as npimport netCDF4 as ncimport cartopy.crs as ccrsds = xr.open_dataset('correlation.1.30.160.200.191.3.13.51.nc')plt.subplot(1,1,1)ds.hgt.plot()plt.show()projection=ccrs.Orthographic(central_latitude=90, central_longitude=0)fig=plt.figure(figsize=(8,8))ax=plt.axes(projection=projection)ax.coastlines()ax.set_global()ax.axhline(0,color='black')ax.axvline(0,color='black')ds=nc.Dataset('correlation.1.30.160.200.191.3.13.51.nc')lon=ds.variables['lon'][:]lat=ds.variables['lat'][:]time_index=0variable=ds.variables['hgt'][time_index,:,:]lonlon,latlat=np.meshgrid(lon,lat)plt.scatter(lonlon,latlat)plt.contourf(lon,lat,variable,cmap='jet')data = ds.variables['time'][:]long = ds.variables['lon'][:]lati = ds.variables['lat'][:]plt.colorbar(label="Sif", orientation="horizontal")cbar=plt.colorbarplt.title('Jan 1948 to 2020: 1000mb Geopotential Height \n Seasonal Correlation w/ Jan SOI \n NCEP /NCAR Reanalysis')plt.show()plt.savefig('12.pdf')给代码改错

data = ds['time'][:] long = ds['lon'][:] lati = ds['lat'][:] plt.colorbar(label="Sif", orientation="horizontal") plt.title('Jan 1948 to 2020: 1000mb Geopotential Height \n Seasonal Correlation w/ Jan ...

'cf_chl_2=; path=/; expires=Thu, Jan 01 1970 00:00:00 UTC; domain=.hapag-lloyd.cn; Secure cf_clearance=_bPJegR0K9V1oLHj1q6CodoEy.VtCIG.5blzf_apuo8-1685521974-0-1-52c890dc.a1562d2.5c9cf3c-150; path=/; expires=Thu, 30-May-24 08:33:17 GMT; domain=.hapag-lloyd.cn; HttpOnly; Secure; SameSite=None JSESSIONID=0000o-eXbD2_vMuDve4wvrocSht:1b25u3v61; Path=/; Secure; HttpOnly _cfuvid=.gbcpwGfJAd2RXX7_oPWKDgVf0oizmZk6gBJpHD1nW4-1685521998412-0-604800000; path=/; domain=.hapag-lloyd.cn; HttpOnly; Secure; SameSite=None' 用js 写正则提取cf_chl_2=; path=

Secure cf_clearance=_bPJegR0K9V1oLHj1q6CodoEy.VtCIG.5blzf_apuo8-1685521974-0-1-52c890dc.a1562d2.5c9cf3c-150; path=/; expires=Thu, 30-May-24 08:33:17 GMT; domain=.hapag-lloyd.cn; HttpOnly; Secure; ...

class args():# training argsepochs = 4 #"number of training epochs, default is 2"batch_size = 4 #"batch size for training, default is 4"dataset = "MSCOCO 2014 path"HEIGHT = 256WIDTH = 256save_model_dir = "models" #"path to folder where trained model will be saved."save_loss_dir = "models/loss" # "path to folder where trained model will be saved."image_size = 256 #"size of training images, default is 256 X 256"cuda = 1 #"set it to 1 for running on GPU, 0 for CPU"seed = 42 #"random seed for training"ssim_weight = [1,10,100,1000,10000]ssim_path = ['1e0', '1e1', '1e2', '1e3', '1e4']lr = 1e-4 #"learning rate, default is 0.001"lr_light = 1e-4 # "learning rate, default is 0.001"log_interval = 5 #"number of images after which the training loss is logged, default is 500"resume = Noneresume_auto_en = Noneresume_auto_de = Noneresume_auto_fn = None# for test Final_cat_epoch_9_Wed_Jan__9_04_16_28_2019_1.0_1.0.modelmodel_path_gray = "./models/densefuse_gray.model"model_path_rgb = "./models/densefuse_rgb.model"

这段代码定义了一个名为args的类，其中包含了许多训练和模型参数的设置。下面是一些重要的参数： - epochs: 训练的轮数，默认为4。 - batch_size: 训练时的批大小，默认为4。 - dataset: 数据集的路径，...

保留原本功能优化以下代码import pandas as pd import numpy as np import matplotlib.pyplot as plt # 1.读取并查看数据 bike_day = pd.read_csv("C:/Users/15020/Desktop/26.bike_day.csv") print(bike_day.head(5)) # 前5行 print(bike_day.tail(2)) #后2行 #2.处理数据并导出到文件 bike_day_user = bike_day[['instant','dteday','yr', 'casual', 'registered']].dropna() bike_day_user.to_csv('bike_day_user.txt', sep=' ',index=False, header=False) #3.读取数据并添加新列并导出到新文件 bike_day_user = pd.read_csv('bike_day_user.txt', sep=' ', header=None, names=['instant','dteday','yr', 'casual',"registered"]) bike_day_user['cnt'] = bike_day_user['casual'] + bike_day_user['registered'] bike_day_user.to_excel('bike_day_user_cnt.xlsx', index=False) #4.读取数据并进行统计 bike_day_user_cnt = pd.read_excel('bike_day_user_cnt.xlsx') print('cnt最大值:',bike_day_user_cnt['cnt'].max()) print('ent最小值:',bike_day_user_cnt['cnt'].min()) print('2011号cnt年平均值:',bike_day_user_cnt[bike_day_user_cnt['yr'] == 0]['cnt'].mean()) print('2012年cnt年平均值:',bike_day_user_cnt[bike_day_user_cnt['yr'] == 1]['cnt'].mean()) print('2011年月严始值:', bike_day_user_cnt[bike_day_user_cnt['yr'] == 0].groupby('mnth')['cnt'].mean()) print('2022年月平均值:', bike_day_user_cnt[bike_day_user_cnt['yr'] == 1].groupby('mnth')['cnt'].mean()) # 5.可视化并保存图像 fig, ax = plt.subplots() ax.barh(bike_day_user_cnt['mnth'], bike_day_user_cnt[bike_day_user_cnt['yr'] == 0].groupby('mnth')['cnt'].mean(), color='blue', label='2011') ax.barh(bike_day_user_cnt['mnth'], bike_day_user_cnt[bike_day_user_cnt['yr'] == 1].groupby('mnth')['cnt'].mean(), color='lightblue', label='2012') ax.set_yticks(np.arange(1,13)) ax.set_yticklabels(['Jan','Feb','Mar', 'Apr', 'May','Jun','Jul','Aug', 'sep', 'Oct','Nov','Dec']) ax.set_xlabel('Average number of shared bike users') ax.set_title('Monthly Average Number of Shared Bike Users in 2011-2012') ax.legend() fig.savefig('bike_day_user_cnt.png', dpi=300)

ax.set_yticklabels(['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']) ax.set_xlabel('Average number of shared bike users') ax.set_title('Monthly Average Number of ...

date_ref_num = datenum('01-jan-1957'); % READING/WRITING THROUGH ALL THE MET_EM DATA %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Initialisation %%%%%%%%%%%%%%%% year_start = str2num( date_start( 1 : 4 ) ); month_start = str2num( date_start( 6 : 7 ) ); day_start = str2num( date_start( 9 : 10 ) ); hour_start = str2num( date_start( 12 : 13 ) ); date_start_num = ( datenum( year_start, month_start, day_start) - ... date_ref_num ) * 24 + hour_start; year_end = str2num( date_end( 1 : 4 ) ); month_end = str2num( date_end( 6 : 7 ) ); day_end = str2num( date_end( 9 : 10 ) ); hour_end = str2num( date_end( 12 : 13 ) ); date_end_num = ( datenum( year_end, month_end, day_end) - ... date_ref_num ) * 24 + hour_end; nb_occurences = ( date_end_num - date_start_num ) / 6 + 1; date_current_num = date_start_num; k_stat = 2; disp(' ')什么意思

这是一段 MATLAB 代码，大概的意思是： - date_ref_num 是一个日期参考值，用于计算时间间隔。 - year_start、month_start、day_start、hour_start 分别是起始时间的年、月、日、时。...

'cf_chl_2=; path=/; expires=Thu, Jan 01 1970 00:00:00 UTC; domain=.hapag-lloyd.cn; Secure cf_clearance=cDrsZgd4k35JCBW9RTlu07QngtEQ5blv5Ki1qgioC50-1685529723-0-1-52c890dc.521115df.5c9cf3c-150; path=/; expires=Thu, 30-May-24 10:42:12 GMT; domain=.hapag-lloyd.cn; HttpOnly; Secure; SameSite=None JSESSIONID=0000MOjix1WTUMkREsY0-do52QV:1b25u3trs; Path=/; Secure; HttpOnly __cf_bm=pQtq2ABXfnlHdahmt31cmoAlkLFALMKt1.MnnOAFXgI-1685529733-0-AQNhOmR/Ihxcdvy858DCc4dj4vNiFXW75bJuXZQyyoodYl1j9FBa2xxcukjBcdcjyUajq0o42KtEholRIgbjaxA=; path=/; expires=Wed, 31-May-23 11:12:13 GMT; domain=.hapag-lloyd.cn; HttpOnly; Secure; SameSite=None _cfuvid=_apIw57_PVdGdqUdVJkHqfi4zCPnI8cOl1cVAg.a0NY-1685529733588-0-604800000; path=/; domain=.hapag-lloyd.cn; HttpOnly; Secure; SameSite=None' js 提取cookie 好的方法是什么

可以使用正则表达式和JavaScript的document.cookie属性来提取cookie。例如，以下代码可以提取cookie中名为cf_clearance的值： var regex = /cf_clearance=([^;]+)/; var match = regex.exec(document.cookie)...

生蚝: （1）用pandas 库读取 PRSA_ data 2010.1.1-2014.12.31.csx 文件，查看前3行、后2行。 (2) 刪除列 DEWP、TEMP、PRES、cbwd、Iws、Is、Ir,并将剰余列用 pandas 数据预处理模块将缺失值丢弃处理，导出到新的csV 文件 pm25_ data 2010.1.1-2014.12.31.csv。 (3）利用pandas 库读取新的数据集 pm25_ data 2010.1.1-2014.12.31.csv，并选择字段 pm2.5 大子300 的所有数据集，昇出力文本文件 pm25_hazardous_data_2010.1.1-2014. 12.31.txt，要求数据之间用逗号分隔，每行末尾包含换行符。 (4）读取文本文件 pm25_hazardous_ data_ 2010.1.1-2014.12.31.txt 并转存到 Excel 文件 pm25_hazardous_data_ 2010.1.1-2014.12.31.xlsx # 生蚝: （5）重新读取文本文件读取文本文件 pm25_ hazardous_ data 2010.1.1-2014.12.31.txt，分别统计出现最多的month、 day、hour，并将 month、 day、hour 的出现频次用柱状图显示。要求包括图例、图标题，xy轴均显示刻度值，柱状图填充颜色分别为红色、绿色、蓝色；并将结果保存为 png 图片保存，分辦率为 400 dpi，png 图片命名分别为 pm25 hazardous_ month_ day hour.png。

1. 使用pandas库读取PRSA_data_2010.1.1-2014.12.31.csv文件，查看前3行、后2行可以使用以下代码： python import pandas as pd data = pd.read_csv('PRSA_data_2010.1.1-2014.12.31.csv') print(data.head(3))...

data_2021 = df[df['中标时间'].str.startswith('2021')] monthly_counts = data_2021.groupby(data_2021['中标时间'].str.slice(5, 7))['标段编号'].count() fig, ax = plt.subplots(figsize=(10, 5)) ax.plot(monthly_counts.index, monthly_counts.values, marker="D") ax.set_xlabel("Month") ax.set_ylabel("Number of Sections") ax.set_title("Monthly Section Counts in 2021") ax.set_xticks(list(range(0, 12))) ax.set_xticklabels(["Jan.", "Feb.", "Mar.", "Apr.", "May", "Jun.", "Jul.", "Aug.", "Sep.", "Oct.", "Nov.", "Dec."]) ax.spines["right"].set_visible(False) ax.spines["top"].set_visible(False) ax.grid(True, axis="y", linestyle="--") for i in range(len(monthly_counts)): ax.text(monthly_counts.index[i], monthly_counts.values[i], f"{monthly_counts.values[i]}", ha="center", va="bottom") ax.plot(monthly_counts.index[i], monthly_counts.values[i], marker="D", markersize=8, color="#D35368") plt.show()

这段代码是用于数据可视化的，主要作用是绘制2021年每个月的标段数量的折线图。首先，通过筛选数据得到2021年的数据；然后，使用groupby方法按月份进行分组，并计算每个月的标段数量；接着，使用matplotlib库绘制...

for line in inputfile_list.readlines(): file_name = line.strip('\n')# 使用strip()函数去掉每行结束的\n if file_name: m1 = re.match("^(\S+)-(\d{2})_(CP\d)-(\S+)$", file_name) # SCWX505A1_C056868.00_C056868.00-02_CP3-RP0_2023JAN15042101_dlogTDO.csv if m1: print (int((m1.group(2))))# 获取匹配的字符，括号里面是分组序号，没有序号则获取整体 wafernum = int((m1.group(2))) m2 = re.match("^(\S+)\.(\S+)$", file_name) # SCWX505A1_C056868.00_C056868.00-04_CP3-RP0_RP1_Merge.xlsx if m2: print (m2.group(1)) file_name_m = m2.group(1)

这段代码是读取一个文件列表，逐行处理每个文件名。首先使用正则表达式匹配文件名中的特定部分，如果匹配成功，则打印出第二个匹配到的分组内容，或者将第二个分组内容转换为整数赋值给变量wafernum。...

优化代码import random import pandas as pd import matplotlib.pyplot as plt fn = 'data.csv' products = ['商品1','商品2','商品3','商品4','商品5','商品6','商品7','商品8','商品9','商品10'] datelist = [] for month in range(1,13) : for day in range(1,32) : date = f'2019-{month:20d}-{day:02d}' datelist.append(date) datalist =[] for date in datelist : for it in products : sales = round(random.uniform(100,1000),2) datalist.append([date,it,sales]) df = pd.DataFrame(datalist, columns=['日期','商品名称','营业额']) df.to_csv('data.csv', index=False) df = pd.read_csv('data.csv') for product in df['products'].unique() : data = df.loc[df['products'] == product] plt.plot(data['date'],data['sales'],label=product) plt.xlabe1('Date') plt.ylabe1('sales') plt.title('Sales by Product') plt.legend() plt.show() df['month'] = pd.DatetimeIndex(df['date']).month groupeddata = df.groupby(['products','month'])['sales'].sum().unstack() groupeddata.plot(kind='bar') plt.xlabel('Products') plt.ylabel('Sales') plt.title('Sales by Month') plt.legend(title='Month',labels=['JAN','FEB','MAR','APR','MAY','JUN','JUL','AUG','SEP','OCT','NOV','DEV']) plt.show() df['quarter'] = pd.PeriodIndex(df['date'],freq='Q') groupeddata = df.groupby(['products','quarter'])['sales'].sum().unstack() groupeddata.plot(kind='pie',subplots=True) plt.title('Sales by Quarter') plt.legend(loc='center left',bbox_to_anchor=(1.0,0.5)) plt.show()

plt.plot(data['日期'],data['营业额'],label=product) plt.xlabel('日期') plt.ylabel('销售额') plt.title('销售额趋势') plt.legend() plt.show() groupeddata = df.groupby(['商品名称','月份'])['营业...

保留原本功能实现的情况下优化以下代码import random import pandas as pd import matplotlib.pyplot as plt fn = 'data.csv' products = ['商品1','商品2','商品3','商品4','商品5','商品6','商品7','商品8','商品9','商品10'] datelist = [] for month in range(1,13) : for day in range(1,32) : date = f'2019-{month:20d}-{day:02d}' datelist.append(date) datalist =[] for date in datelist : for it in products : sales = round(random.uniform(100,1000),2) datalist.append([date,it,sales]) df = pd.DataFrame(datalist, columns=['日期','商品名称','营业额']) df.to_csv('data.csv', index=False) df = pd.read_csv('data.csv') for product in df['products'].unique() : data = df.loc[df['products'] == product] plt.plot(data['date'],data['sales'],label=product) plt.xlabe1('Date') plt.ylabe1('sales') plt.title('Sales by Product') plt.legend() plt.show() df['month'] = pd.DatetimeIndex(df['date']).month groupeddata = df.groupby(['products','month'])['sales'].sum().unstack() groupeddata.plot(kind='bar') plt.xlabel('Products') plt.ylabel('Sales') plt.title('Sales by Month') plt.legend(title='Month',labels=['JAN','FEB','MAR','APR','MAY','JUN','JUL','AUG','SEP','OCT','NOV','DEV']) plt.show() df['quarter'] = pd.PeriodIndex(df['date'],freq='Q') groupeddata = df.groupby(['products','quarter'])['sales'].sum().unstack() groupeddata.plot(kind='pie',subplots=True) plt.title('Sales by Quarter') plt.legend(loc='center left',bbox_to_anchor=(1.0,0.5)) plt.show()

def generate_data(): fn = 'data.csv' products = ['商品1','商品2','商品3','商品4','商品5','商品6','商品7','商品8','商品9','商品10'] datelist = [] for month in range(1,13): for day in range(1,32): ...

这个错误是由于函数 get_nb2_ip() 没有返回合法的响应导致的。在这个函数中，需要返回一个字符串、字典、列表、元组（带有头部或状态）、Response 实例或 WSGI 可调用对象。可以尝试检查这个函数的代码并确保它返回...

如果有 months = "Jan.Feb.Mar.Apr.May.Jun.Jul.Aug.Sep.Oct.Nov.Dec."，编写程序，用户输入月份数字，输出月份缩写

months = "Jan.Feb.Mar.Apr.May.Jun.Jul.Aug.Sep.Oct.Nov.Dec." month_num = int(input("请输入月份数字：")) if month_num >= 1 and month_num <= 12: start_index = (month_num - 1) * 4 end_index = start_...

JAN_mean = data_select.mean('time') - CSDN文库 (2024)

相关推荐

References