条形图.qmd

# 条形图

```{r}
library(tidyverse)
library(ggpubr)
library(ggthemes)
library(patchwork)
```

## 修改边框和填充颜色

```{r}
df <- ToothGrowth |> rownames_to_column("id")
ggplot(df,aes(dose,len,color=factor(dose)))+
    geom_col(fill=NA)
```


## 添加标签

```{r}
ggplot(df,aes(id,len))+
    geom_col()+
   ggrepel::geom_text_repel(aes(y=len+3,label=len))
```

## 添加误差棒

```{r}
df <- ToothGrowth |> rownames_to_column("id")
ggplot(df, aes(x=supp, y=len, fill = factor(dose)))+
  stat_summary(fun="mean", geom="bar",position = position_dodge(1))+
  stat_summary(fun.data="mean_sd", geom="errorbar", width=0.25,position = position_dodge(1))+
  theme_tufte(base_size=16, base_family="sans", ticks=FALSE)+
  labs(x="Group", y="Values", caption="Error Bars Indicate Standard Error")
```

## 添加显著性

```{r}
library(ggsignif)
ggplot(data = PlantGrowth, aes(x = group, y = weight)) +
  geom_boxplot(width = 0.25) +
  geom_jitter(width = 0.15) +
  geom_signif(comparisons = list(c("ctrl", "trt1"), c("trt1", "trt2")), 
              map_signif_level = function(p) sprintf("p = %.3g", p), 
              textsize = 6, test = "t.test") +
    ylim(c(NA,7))+
  theme_classic() +
  coord_cartesian(clip = "off")
```

## 添加散点、抖动点

```{r}
ggplot(mpg, aes(x = drv, y = hwy, fill = as.factor(cyl))) +  
    geom_bar(stat = "summary",fun=mean, position = position_dodge(1)) +
    stat_summary(fun.data = 'mean_sd',geom="errorbar",position = position_dodge(1),
                 width=.2,linewidth=1)+
    geom_point(position = position_dodge(1), 
              size = 3,alpha=0.3
             )  
```

```{r}
df <- ToothGrowth
ggplot(df, aes(x = supp, y = len)) +
    geom_bar(
        aes(fill = factor(dose)),
        stat = "summary",
        fun = mean,
        position = position_dodge(width = 0.9)
    ) +
    geom_point(position = position_jitterdodge(jitter.width = 0.1),
               aes(color = factor(dose)),
               alpha = 0.7)+
    scale_color_manual(values = c("red","green","blue"))|
    ggplot(df, aes(x = supp, y = len)) +
    geom_bar(
        aes(fill = factor(dose)),
        stat = "summary",
        fun = mean,
        position = position_dodge(width = 0.9)
    ) +
    geom_dotplot(position = position_dodge(width = 0.9),
               aes(fill = factor(dose)),dotsize = 0.8,
               binaxis = "y", stackdir = "center",color="black",
               alpha = 0.7)+
    scale_fill_manual(values = c("red","green","blue"))
```

## 分组排序条图 waterfall barplot

```{r}
library(tidyverse)
library(ggpubr)
library(ggsci)
library(readxl)
c <- read_excel("data/01source.xlsx",sheet=1,range = "A2:D16") |> 
  rename(
    LR_DPVB=`LR-DPVB`
  )  |> 
  pivot_longer(
    cols = CON:LR_DPVB,
    names_to = "treatment",
    values_to = "changerate"
  ) |> 
  mutate(
    treatment=factor(treatment,levels=c('CON','LDRT','DPVB','LR_DPVB')),
    changerate=if_else(changerate>=1,100,changerate*100)
  ) |> 
  arrange(treatment)|>
  drop_na() |> 
  mutate(
    id=1:52
    ) 
```

```{r}

ggplot( c ,aes(x=id,y=changerate,fill=treatment))+ 
  geom_bar(stat = "identity",position = "dodge",width =0.8)+
    scale_fill_jco()+
    scale_y_continuous(name = "MRI tumor volume change from baseline (%)",
                       limits = c(-100,100),
                       breaks = seq(-100,100,50),
                       )+
    theme_pubr()+
    theme(
        legend.position = c(0.8,0.8),
        legend.title = element_blank(),
        axis.title.x = element_blank(),
        axis.text.x = element_blank(),
        axis.ticks.x = element_blank(),
        axis.line.x = element_blank(),
        plot.title = element_text(hjust = 0.5),
    )+
    ggtitle("Hepa1-6 tumor")+
    geom_hline(yintercept = 0,color = "black",linetype =1,linewidth=0.8)+
    geom_hline(yintercept=c(20,-30), colour="gray15", linetype=3,linewidth=0.8)+                # 0 = blank, 1 = solid, 2 = dashed, 3 = dotted, 
            # 4 = dotdash, 5 = longdash, 6 = twodash
    annotate(geom = "text",x=c(2,50),y=c(-30,20),
             label = c("PR:-30%","PD:20%"),
             vjust = -0.5,
             )   

```

```{r}

library(ggplot2)
library(rstatix)
library(ggpubr)
# sample data
plot_df <- ToothGrowth

# Summarize the data
summary_df <- plot_df %>%
  group_by(dose, supp) %>%
  summarise(
    Mean = mean(len),
    SD = sd(len),
    SE = sd(len) / sqrt(n())  # Standard Error
  )

# p values for supp between dose
p_sup <- plot_df %>%
  group_by(supp) %>%
  t_test(len ~ dose, p.adjust.method = "fdr")

p_sup <- p_sup %>% 
  add_xy_position(x = "dose", group = "supp")

# Group the data by dose and then compare the levels of the supps variable
p_dose <- plot_df %>%
  group_by(dose) %>%
  t_test(len ~ supp) %>%
  adjust_pvalue(method = "fdr") %>%
  add_significance("p.adj")
p_dose

# Add p-values onto the box plots
p_dose <- p_dose %>%
  add_xy_position(x = "dose")

# change dose to character
summary_df$dose <- as.character(summary_df$dose)




ggplot(summary_df, aes(x = dose, y = Mean, fill = supp)) +
  geom_bar(stat = "identity", color = "black", position = position_dodge()) +
  geom_errorbar(aes(ymin = Mean - SE, ymax = Mean + SE), width = 0.2, position = position_dodge(.9)) +
  theme_minimal() +
  scale_fill_manual(
    values = c("#999999", "#E69F00"),
    aesthetics = c("color", "fill")
  ) +
  labs(
    title = "ToothGrowth",
    x = "dose",
    y = "len ± SE"
  ) +
  theme(
    plot.title = element_text(hjust = 0.5)
  ) +
  stat_pvalue_manual(
    p_sup,
    tip.length = 0.01,
    hide.ns = F,
    color = "supp"
  ) +
  stat_pvalue_manual(
    p_dose,
    tip.length = 0.01,
    hide.ns = F,
    inherit.aes = FALSE
  )
```

## 复杂条图

### 分组条形图添加 errorbar、significance 、dotplot

[ggpubr](https://www.datanovia.com/en/blog/author/kassambara/)

```{r}
g<-read_excel("data/01source.xlsx",sheet=1,range="G2:Z3") 
colnames(g) <- rep(c('CON','LDRT','DPVB','LR_DPVB'),each=5)
g<- g |>  pivot_longer(
  cols = everything(),
  names_to = "treatment",
  values_to = "liverweight"
  )|> 
  mutate(
    treatment=factor(treatment,levels=c('CON','LDRT','DPVB','LR_DPVB'))
  )
g
```

```{r}
source("function/calculate_t_tests.R")
calculate_t_tests(g,"treatment","liverweight")
```

::: callout-note
函数calculate_t_tests

```{r}
#| label: calculate_t_tests()
# 定义计算两两组之间t检验的函数
calculate_t_tests <- function(data, group_by, value_column) {
  # data: 数据框
  # group_by: 组别的列名
  # value_column: 要比较的数值的列名
  
  # 提取唯一的组别
  groups <- unique(data[[group_by]])
  
  # 初始化结果矩阵
  result_matrix <- matrix(NA, nrow = length(groups), ncol = length(groups),
                          dimnames = list(groups, groups))
  
  # 循环遍历所有可能的组合
  for (i in 1:(length(groups)-1)) {
    for (j in (i+1):length(groups)) {
      # 提取两组数据
      group1_data <- data[data[[group_by]] == groups[i], value_column]
      group2_data <- data[data[[group_by]] == groups[j], value_column]
      
      # 执行t检验
      t_test_result <- t.test(group1_data, group2_data)
      
      # 提取p值
      p_value <- t_test_result$p.value
      
      # 将p值存入结果矩阵
      #result_matrix[groups[i], groups[j]] <- p_value
      result_matrix[groups[j], groups[i]] <- p_value
    }
  }
  
  # 返回结果矩阵
  return(result_matrix)
}

```
:::

### ggpubr

```{r}
library(ggpubr)
ggbarplot(data = g,
          x="treatment",y="liverweight",
          add = c("mean_sd"),
          fill = "treatment",
          palette = "jco",
          title = "DEN+CCl4 tumor",
          xlab = "",
          ylab='liver weight (g)',
          legend='none',
          )+
  stat_compare_means(
      #aes(label='p.format'),
      comparisons = list(c('CON','LDRT'),c("DPVB",'LR_DPVB'), c("CON","LR_DPVB")),
      method = 't.test',
      tip.length = c(0,0,0,0,0,0),
      bracket.size = 1,linewidth=1)+

  theme(plot.title = element_text(hjust = 0.5),
        axis.text.x = element_text(angle = 45,vjust = 0.5))+
  geom_dotplot(data = g,
               mapping=aes(x=treatment,y=liverweight),
               binaxis = 'y',stackdir = 'center',binwidth = 0.05)

```

### ggplot2+ggsignif

```{r}
library(ggsignif)
g |> 
  summarise(
    n=n(),
    mean=mean(liverweight),
    sd=sd(liverweight),
    .by=treatment
  ) |> 
    ggplot(aes(treatment,mean,fill=treatment))+
    geom_bar(stat = "identity",width = .7,position = position_dodge())+#条形图
    geom_errorbar(aes(ymin=mean-sd,ymax=mean+sd),
                  width=.2,color="black",linewidth=.8)+#误差棒
    geom_signif(data = g,
              mapping=aes(x=treatment,y=liverweight),
              comparisons = list(c('CON','LDRT'),c("DPVB",'LR_DPVB'),
                                 c("CON","LR_DPVB")),
              annotations = c("p=0.8639","p=0.0088","p=0.00034"),
              map_signif_level=F,
              tip_length=c(0.0,0.0,0.0,0.0,0.0,0.0),
              y_position=c(2.1,1.8,2.35),size = .8,
              textsize = 5,
              test = "t.test"
              )+
    theme_pubr()+
    scale_fill_jco()+
    scale_y_continuous(limits = c(0,2.5))+
    labs(
        x='',
        y='liver weight (g)',
        title="DEN+CCl4 tumor"
    )+
    theme(
        plot.title = element_text(hjust =0.5 ),
        legend.position = 'none',
        axis.text.x = element_text(angle = 45,vjust = .5)
    )+  
    geom_dotplot(data = g,
                 mapping=aes(x=treatment,y=liverweight),
                 binaxis = 'y',
                 stackdir = 'center',#居中
                 fill='black',
                 binwidth = 0.05)
  
```

### 分组并列，组内显著性p值

<https://www.datanovia.com/en/blog/tag/ggpubr/>

<https://www.datanovia.com/en/blog/how-to-add-p-values-onto-a-grouped-ggplot-using-the-ggpubr-r-package/>

```{r}
f2e <- read_excel("data/01source.xlsx", sheet = 3, range = "A2:M4")
colnames(f2e) <-
    c("type", rep(c('CON', 'LDRT', 'DPVB', 'LR_DPVB'), each = 3))

f2e |> pivot_longer(cols = -1,
                    names_to = "trt",
                    values_to = "pct",
) |>
    mutate(trt = factor(trt, levels = c('CON', 'LDRT', 'DPVB', 'LR_DPVB'))) -> f2e
```

```{r}
library(ggpubr)
library(rstatix)
 conflicts_prefer(rstatix::t_test)
stat.test <- f2e %>%
  group_by(type) %>%
  t_test(pct ~ trt) |> 
  adjust_pvalue(method = "bonferroni") %>%
    add_significance("p.adj") |> 
    add_significance("p") 


stat.test <- stat.test %>%
  add_xy_position(fun = "mean_sd", x = "type", dodge = 1)|> 
  dplyr::filter(group1=="CON"&group2=="LR_DPVB")


e_left <- ggbarplot(f2e, x = "type", y = "pct",
                fill = "trt", palette = "jco",
                add =c("mean_sd"), add.params = list(group = "trt"),
                position = position_dodge(1),
                #legend="none",
          legend.title="",
          )+
    scale_y_continuous(limits = c(0,80))+
    stat_pvalue_manual(
        data = stat.test,
        label = "p={p}({p.signif})",
        tip.length = 0.05,
        y.position = c(40,75),
        #bracket.nudge.y = -2,
)+labs(
        x='',
        y='% in live CD45+ cell',
        #title="Hepa1-6 tumor"
    )

e_right <- e_left

library(patchwork)
(e_left + e_right)+
    plot_annotation(
        title = "Hepa1-6 tumor",
    ) +
    plot_layout(guides = "collect")& 
    theme(
        plot.title = element_text(hjust = 0.5),
        legend.position = "top",
    )
```

### 分组并列，组内组间显著性p值

<https://www.datanovia.com/en/blog/how-to-add-p-values-onto-a-grouped-ggplot-using-the-ggpubr-r-package/>

```{r}
library(ggpubr)
library(rstatix)
df <- ToothGrowth
df$dose <- as.factor(df$dose)



bp <- ggbarplot(
    df,
    x = "dose",
    y = "len",
    add = "mean_sd",
    color = "supp",
    palette = c("#00AFBB", "#E7B800"),
    position = position_dodge(0.8)
)


# 统计检验
stat.test <- df %>%
    group_by(dose) %>%
    t_test(len ~ supp) %>%
    adjust_pvalue(method = "bonferroni") %>%
    add_significance("p.adj")
stat.test <- stat.test %>%
    add_xy_position(fun = "mean_sd", x = "dose", dodge = 0.8)


stat.test2 <- df %>%
    t_test(len ~ dose, p.adjust.method = "bonferroni")
stat.test2 <- stat.test2 %>%
    add_xy_position(x = "dose")

pwc <- df %>%
    group_by(supp) %>%
    t_test(len ~ dose, p.adjust.method = "bonferroni")
pwc
pwc <- pwc %>%
    add_xy_position(
        x = "dose",
        fun = "mean_sd",
        group = "supp",
        dodge = 0.8
    )

# 组间组内复合


bp +
     stat_pvalue_manual(stat.test,  label = "p.adj.signif", tip.length = 0.01) 
    # stat_pvalue_manual(
    #     stat.test2,
    #     label = "p",
    #     tip.length = 0.02,
    #     step.increase = 0.05,
    # ) +
    # scale_y_continuous(expand = expansion(mult = c(0.05, 0.1))) +
    # stat_pvalue_manual(
    #     pwc,
    #     color = "supp",
    #     step.group.by = "supp",
    #     tip.length = 0,
    #     step.increase = 0.1,
    # )
```