eda_lazyman.Rmd

# 懒人系列 {#eda-lazyman}


R社区上很多大神，贡献了很多非常优秀的工具，节省了我们的时间，也给我们的生活增添了无限乐趣。我平时逛github的时候时整理一些，现在分享出来供像我一样的懒人用，因此本文档叫“懒人系列”。欢迎大家补充。

## 列名太乱了
```{r lazyman01, message = FALSE, warning = FALSE}
library(tidyverse)
library(janitor)
## install.packages("janitor")
## https://github.com/sfirke/janitor
```

```{r lazyman-1}
fake_raw <- tibble::tribble(
  ~id, ~`count/num`, ~W.t, ~Case, ~`time--d`, ~`%percent`,
  1L, "china", 3L, "w", 5L, 25L,
  2L, "us", 4L, "f", 6L, 34L,
  3L, "india", 5L, "q", 8L, 78L
)
fake_raw
```


```{r lazyman-2}
fake_raw %>% janitor::clean_names()
```


## 比count()更懂我的心
```{r lazyman-3}
mtcars %>%
  dplyr::count(cyl, sort = TRUE) %>% 
  mutate(percent = 100 * n / sum(n))
```

```{r lazyman-4, eval=FALSE}
mtcars %>%
  janitor::tabyl(cyl)
```



## 比distinct()更知我心

```{r lazyman-5}
df <- tribble(
  ~id, ~date, ~store_id, ~sales,
  1, "2020-03-01", 1, 100,
  2, "2020-03-01", 2, 100,
  3, "2020-03-01", 3, 150,
  4, "2020-03-02", 1, 110,
  5, "2020-03-02", 3, 101
)

df %>%
  janitor::get_dupes(store_id)

df %>%
  janitor::get_dupes(date)
```




## 代码太乱了，谁帮我整理下

```{r lazyman-6}
## install.packages("styler")
```


```{r lazyman-7, out.width = '100%', echo = FALSE}
knitr::include_graphics("images/styler.png")
```

安装后，然后这两个地方点两下，就发现你的代码整齐很多了。或者直接输入

```{r lazyman-8, eval=FALSE}
styler:::style_active_file()
```

## 用datapasta粘贴小表格

有时候想把excel或者网页上的小表格，放到R里测试下，如果用readr读取excel小数据，可能觉得麻烦，或者大材小用。比如网页<https://en.wikipedia.org/wiki/Table_(information)>有个表格，有偷懒的办法弄到R？

```{r, out.width = '100%', echo = FALSE}
knitr::include_graphics("images/datapaste1.png")
```


推荐一个[方法](https://github.com/MilesMcBain/datapasta)

1、安装`install.packages("datapasta")`
2、鼠标选中并复制网页中的表格
3、在 Rstudio 中的`Addins`找到`datapasta`，并点击`paste as tribble`

```{r, out.width = '100%', echo = FALSE}
knitr::include_graphics("images/datapaste2.png")
```

## 谁帮我敲模型的公式

```{r lazyman-9}
library(equatiomatic)
## https://github.com/datalorax/equatiomatic
```


```{r lazyman-10}
mod1 <- lm(mpg ~ cyl + disp, mtcars)
```


```{r lazyman-11, results="asis"}
extract_eq(mod1)
```


```{r lazyman-12, results="asis"}
extract_eq(mod1, use_coefs = TRUE)
```

## 模型有了，不知道怎么写论文？

```{r lazyman-13}
library(report)
## https://github.com/easystats/report
```


```{r lazyman-report, results="asis"}
model <- lm(Sepal.Length ~ Species, data = iris)
report(model)
```


## 模型评估一步到位
```{r lazyman-performance}
library(performance)

model <- lm(mpg ~ wt * cyl + gear, data = mtcars)
performance::check_model(model)
```



## 统计表格不用愁
```{r lazyman-gtsummary1, results="asis", eval=FALSE}
library(gtsummary)
## https://github.com/ddsjoberg/gtsummary


gtsummary::trial %>%
  dplyr::select(trt, age, grade, response) %>%
  gtsummary::tbl_summary(
    by = trt,
    missing = "no"
  ) %>%
  gtsummary::add_p() %>%
  gtsummary::add_overall() %>%
  gtsummary::add_n() %>%
  gtsummary::bold_labels()
```


直接复制到论文即可
```{r lazyman-gtsummary2, results="asis", eval=FALSE}
t1 <-
  glm(response ~ trt + age + grade, trial, family = binomial) %>%
  gtsummary::tbl_regression(exponentiate = TRUE)

t2 <-
  survival::coxph(survival::Surv(ttdeath, death) ~ trt + grade + age, trial) %>%
  gtsummary::tbl_regression(exponentiate = TRUE)



gtsummary::tbl_merge(
  tbls = list(t1, t2),
  tab_spanner = c("**Tumor Response**", "**Time to Death**")
)
```



## 统计结果写图上

```{r lazyman-statsExpressions, out.width='99%', eval=FALSE, include=TRUE}
library(ggplot2)
library(statsExpressions)
# https://github.com/IndrajeetPatil/statsExpressions


ggplot(mtcars, aes(x = mpg, y = wt)) +
  geom_point() +
  geom_smooth(method = "lm") +
  labs(
    title = "Spearman's rank correlation coefficient",
    subtitle = expr_corr_test(mtcars, mpg, wt, type = "nonparametric")
  )
```




## 正则表达式太南了

```{r lazyman-14}
library(inferregex)
## remotes::install_github("daranzolin/inferregex")
```


```{r lazyman-15}
s <- "abcd-9999-ab9"
infer_regex(s)$regex
```

有了它，妈妈再也不担心我的正则表达式了

## 颜控怎么配色？

```{r lazyman-16}
library(ggthemr) ## devtools::install_github('cttobin/ggthemr')
ggthemr("dust")
```



```{r lazyman-17}
mtcars %>%
  mutate(cyl = factor(cyl)) %>%
  ggplot(aes(x = mpg, fill = cyl, colour = cyl)) +
  geom_density(alpha = 0.75) +
  labs(fill = "Cylinders", colour = "Cylinders", x = "MPG", y = "Density") +
  legend_top()
```

用完别忘了
```{r lazyman-18}
ggthemr_reset()
```


## 画图颜色好看不
scales也是大神的作品，功能多多
```{r lazyman-19}
## https://github.com/r-lib/scales
library(scales)

show_col(viridis_pal()(10))
```

不推荐个人配色，因为我们不专业。直接用专业的配色网站
[colorbrewer](https://colorbrewer2.org/)

先看看颜色，再选择


## 宏包太多
```{r lazyman-20, eval=FALSE}
library(pacman)
## p_load(lattice, foreign, boot, rpart)
```


唉，这个`library()`都要偷懒，真服了你们了


## 犹抱琵琶半遮面
```{r lazyman-gganonymize}
## https://github.com/EmilHvitfeldt/gganonymize
library(ggplot2)
library(gganonymize)

ggg <-
  ggplot(mtcars, aes(as.factor(cyl))) +
  geom_bar() +
  labs(
    title = "Test title",
    subtitle = "Test subtitle, this one have a lot lot lot lot lot more text then the rest",
    caption = "Test caption",
    tag = 1
  ) +
  facet_wrap(~vs)

gganonomize(ggg)
```


你可以看我的图，但就不想告诉你图什么意思，因为我加密了


## 整理Rmarkdown
```{r lazyman-21}
# remotes::install_github("tjmahr/WrapRmd")
# remotes::install_github("fkeck/quickview")
# remotes::install_github("mwip/beautifyR")
```



## 如何有效的提问
直接看官方网站，这里不举例了
```{r lazyman-22}
## install.packages("reprex")
## https://reprex.tidyverse.org/
```

## 程序结束后记得提醒我
```{r lazyman-23}
## beepr::beep(sound = "mario")
```

你听到了声音吗?

## 多张图摆放
```{r lazyman-patchwork}
library(patchwork)
p1 <- ggplot(mtcars) +
  geom_point(aes(mpg, disp))

p2 <- ggplot(mtcars) +
  geom_boxplot(aes(gear, disp, group = gear))

p3 <- ggplot(mtcars) +
  geom_smooth(aes(disp, qsec))

p1 + p2 + p3
```

## 缺失值处理
```{r lazyman-naniar}
library(naniar)
## https://github.com/njtierney/naniar

airquality %>%
  group_by(Month) %>%
  naniar::miss_var_summary()
```


## 看看数据什么情况
```{r lazyman-24}
library(visdat)

vis_dat(airquality)
```






## 管道都不想
管道都不想写， 写代码还有美感？
```{r lazyman-25}
## library(nakepipe)
```


## 各种插件，任君选取
```{r lazyman-26}
## https://github.com/daattali/addinslist
```





```{r lazyman-27, echo = F}
# remove the objects
# rm(list=ls())
rm(df, fake_raw, ggg, mod1, model, p1, p2, p3, s)
```

```{r lazyman-28, echo = F, message = F, warning = F, results = "hide"}
pacman::p_unload(pacman::p_loaded(), character.only = TRUE)
```