R: R CSV 文件读写：readr 完整教程

id,date,product,region,amount
1,2024-01-15,手机,北京,2999
2,2024-01-15,电脑,上海,5999
3,2024-01-16,手机,广州,2899
...
（@,000 rows of）

# 1. 用 readr 智能读取（5 秒）
library(readr)
sales <- read_csv("sales.csv")

# 2. 自动类型推断
# id → integer, date → date, product → character,
# region → character, amount → double

# 3. 只读取需要的行（提速 3 倍）
sales_sample <- read_csv("sales.csv", n_max = 10000)

# 4. 写出干净报表
write_csv(sales_sample, "sales_clean.csv")

姓名,年龄,城市
Alice,25,北京
Bob,30,上海
Charlie,35,广州

# 基础 R（老代码常见）
df_old <- read.csv("data.csv")
# 默认行为：所有字符串列转 factor（坑！）

# tidyverse（推荐）
library(readr)
df_new <- read_csv("data.csv")
# 智能推断类型，字符保持字符

read_csv(file, col_types = TRUE, col_names = TRUE, na = "NA",
         skip = 0, n_max = Inf, locale = default_locale(),
         progress = TRUE)

library(readr)

# 1. 基本读取（智能推断类型）
df <- read_csv("data.csv")

# 2. 指定列类型（避免推断错误）
df <- read_csv("data.csv", col_types = cols(
  id = col_integer(),
  date = col_date(format = "%Y-%m-%d"),
  amount = col_double(),
  name = col_character()
))

# 3. 跳过前 5 行注释
df <- read_csv("data.csv", skip = 5)

# 4. 自定义 NA 标记（CSV 里用 "NULL" 表示缺失）
df <- read_csv("data.csv", na = c("", "NA", "NULL", "N/A"))

# 5. 只读前 1000 行（调试用）
df <- read_csv("data.csv", n_max = 1000)

# 6. 跳过列（用 col_types 设为 col_skip()）
df <- read_csv("data.csv", col_types = cols(
  temp_col = col_skip()
))

# 推断示例
sales <- read_csv("sales.csv")
spec(sales)  # 查看推断结果
# cols(
#   id = col_integer(),
#   date = col_date(format = "%Y-%m-%d"),
#   product = col_character(),
#   region = col_character(),
#   amount = col_double()
# )

write_csv(x, file, na = "NA", append = FALSE, col_names = TRUE)

# 1. 基本输出
write_csv(sales, "sales_clean.csv")

# 2. 输出到指定路径
write_csv(sales, "output/2024_sales.csv")

# 3. 追加到现有文件（不覆盖）
write_csv(new_data, "sales.csv", append = TRUE, col_names = FALSE)

# 4. 自定义 NA 表示
write_csv(df, "output.csv", na = "")

# 写出
write_csv(sales, "sales.csv")

# 再读回
sales_loaded <- read_csv("sales.csv")
# 类型由 read_csv 重新推断

# 1. 指定编码读取
df <- read_csv("data.csv", locale = locale(encoding = "UTF-8"))
# 或
df <- read_csv("data.csv", locale = locale(encoding = "GBK"))

# 2. 检查文件编码
guess_encoding("data.csv")
# 多数情况返回 "UTF-8" 或 "GBK"

# 3. 写入时指定编码
write_csv(df, "output.csv")  # 默认 UTF-8

# 1. 制表符分隔（TSV）
df <- read_tsv("data.tsv")  # sep = "\t"

# 2. 分号分隔（欧洲常用）
df <- read_csv2("data.csv")  # sep = ";"

# 3. 自定义分隔符（如 |）
df <- read_delim("data.txt", delim = "|")

# 欧洲格式：1.234,56（千分位是 .，小数是 ,）
df <- read_csv("data_eu.csv",
               locale = locale(grouping_mark = ".", decimal_mark = ","))

# 1. 只读需要的列（提速 2-3 倍）
df <- read_csv("big.csv", col_select = c(id, amount))

# 2. 用 col_types 跳过列
df <- read_csv("big.csv", col_types = cols(
  temp_col_1 = col_skip(),
  temp_col_2 = col_skip()
))

# 3. 关闭进度条
df <- read_csv("big.csv", progress = FALSE)

# 4. 更快的方案：data.table::fread
# fread("big.csv") 比 read_csv 快 5-10 倍

col_types = cols(
  id = col_integer(),       # 整数
  name = col_character(),   # 字符
  amount = col_double(),    # 双精度浮点
  date = col_date(),        # 日期
  time = col_time(),        # 时间
  datetime = col_datetime(),# 日期时间
  is_active = col_logical(),# 逻辑
  skip = col_skip(),        # 跳过该列
  guess = col_guess()       # 让 readr 推断
)

# 用 "i" "c" "d" "D" "l" 简写
col_types = cols(
  id = "i",          # integer
  name = "c",        # character
  amount = "d",      # double
  date = "D",        # date
  is_active = "l"    # logical
)

# 全部推断（默认）
col_types = TRUE

# 用列表（顺序对应列）
col_types = list(
  id = col_integer(),
  name = col_character()
)

# ============================================
# 销售数据导入与处理
# 功能：读取 CSV → 清洗 → 分析 → 输出报表
# ============================================

library(readr)
library(dplyr)

# 1. 准备示例数据
sample_data <- tibble(
  id = 1:10,
  date = as.Date("2024-01-01") + 0:9,
  product = c("手机", "电脑", "手机", "平板", "手机",
              "电脑", "平板", "手机", "电脑", "手机"),
  region = c("北京", "上海", "广州", "北京", "深圳",
            "上海", "广州", "北京", "深圳", "上海"),
  amount = c(2999, 5999, 2899, 3999, 3099,
            6299, 3899, 2899, 5999, 3199),
  note = c(NA, "促销", NA, NA, "新品", NA, NA, "促销", NA, NA)
)

# 写入示例 CSV（UTF-8 编码）
write_csv(sample_data, "sales_demo.csv")

cat("=== 文件已生成 ===\n")

# 2. 读取 CSV（智能推断）
sales <- read_csv("sales_demo.csv",
                  na = c("", "NA"),
                  col_types = cols(
                    id = col_integer(),
                    date = col_date(format = "%Y-%m-%d"),
                    product = col_character(),
                    region = col_character(),
                    amount = col_double(),
                    note = col_character()
                  ))

cat("=== 读取结果 ===\n")
print(sales)

# 3. 数据质量检查
cat("\n=== 数据结构 ===\n")
str(sales)

# 4. 数据分析
cat("\n=== 销售统计 ===\n")
summary_stats <- sales |>
  group_by(region) |>
  summarise(
    订单数 = n(),
    总销售 = sum(amount),
    平均订单 = round(mean(amount), 2),
    最高订单 = max(amount)
  ) |>
  arrange(desc(总销售))

print(summary_stats)

# 5. 找出促销订单
cat("\n=== 促销订单 ===\n")
promo <- sales |> filter(!is.na(note))
print(promo)

# 6. 按产品分类
cat("\n=== 产品销售 ===\n")
product_stats <- sales |>
  group_by(product) |>
  summarise(销量 = n(), 销售额 = sum(amount))
print(product_stats)

# 7. 输出干净报表
write_csv(summary_stats, "sales_by_region.csv")
write_csv(product_stats, "sales_by_product.csv")
cat("\n=== 报表已输出 ===\n")
cat("  - sales_by_region.csv（按地区）\n")
cat("  - sales_by_product.csv（按产品）\n")

# 8. 用 RDS 保留类型（推荐）
saveRDS(sales, "sales_clean.rds")
sales_reloaded <- readRDS("sales_clean.rds")
cat("\nRDS 加载后类型：\n")
print(sapply(sales_reloaded, class))

=== 销售统计 ===
# A tibble: 4 × 5
  region 订单数 总销售 平均订单 最高订单
  <chr>   <int>   <dbl>    <dbl>    <dbl>
1 北京       3   9897    3299     3999
2 上海       3  15497    5166.    6299
3 广州       2   6798    3399     3899
4 深圳       2   9098    4549.    5999

=== 促销订单 ===
# A tibble: 2 × 6
     id date       product region amount note
  <int> <date>     <chr>   <chr>   <dbl> <chr>
1     2 2024-01-02 电脑    上海     5999 促销
2     8 2024-01-08 手机    北京     2899 促销