Skip to main content

Community Contributions STAT GR 5293

39 class_material_code_throughout_this_semester

Weijia Wang

Parameters:

shape: changes the shape of points into squares, triangles according to the category of variable

color: changes colors, specify types.

size: changes the size of shapes?

scale_fill_manual(values = c(“#ff99ff”, “#cc9966”)) # fill color into bar chart

span: geom_smooth(method = “loess”, span = .1, se = FALSE)

39.1 Continuous variables

### Histograms 1) base r: hist(x, col = “color”) 2) ggplot2: geom_histogram(color = “blue”, fill = “lightblue”, binwidth = 5, center = 52.5) + theme_grey(14))

### Bindwidth 1) bin boundaries right closed (55, 60] or right open [55, 60) code: right = TRUE or FALSE hist(x, right = FALSE, ylim = c(0, 4), xlab = “right open ex. [55, 60)”, font.lab = 2) 2) set bin boundaries by hand: code: breaks = seq(start, stop, gap) hist(x, breaks = seq(47.5, 72.5, 5), col = “lightblue”, axes = FALSE) 3) Frequency (count) histogram hist(prices, breaks = seq(300, 800, 100), col = “lightblue”, las = 1) 4) Density histogram: freq = FALSE hist(prices, breaks = c(300, 400, 500, 600, 700, 800), freq = FALSE, col = “lightblue”, ylab = ““, main =”Density Histogram”, las = 1) 5) Calculate: Density histogram with unequal bin widths g2 <- ggplot(df, aes(x = center, y = percent/(100*binwidth), width = binwidth)) + geom_col(color = “blue”, fill = “lightblue”) + ylab(“density”) + xlab(“age”) + scale_x_continuous(breaks = c(0, df$breaks)) + ggtitle(“Census 2000: Zip Code 10027”) 6) Cumulative frequency histogram: geom_histogram(aes(y = cumsum(..count..)), color = “blue”, fill = “lightblue”) 7) Change bindwidth interactively: ggvis df %>% ggvis(~GDP) %>% layer_histograms(fill := “green”, width = input_slider(500, 10000, value = 5000, step = 500, label = “width”))

39.2 Boxplots

### How to draw: ggplot(tidySavings, aes(person, amount)) + geom_boxplot() ### Add number to quantiles: boxplot(D, horizontal = TRUE, ylim=c(-250, 200)) text(fivenum(D)[c(1,3,5)], 1.25, round(fivenum(D)[c(1,3,5)],1), col = “red”) text(fivenum(D)[c(2,4)], .75, round(fivenum(D),1)[c(2,4)], col = “red”) ### Outliers: if the observation is 2.5 times further than the lower hinge, then it should be an outlier ### Variable width boxplots: geom_boxplot(varwidth = TRUE) ### Multiple density histograms, ordered by median gh <- ggplot(world, aes(x = TFR, y = ..density..)) + geom_histogram(color = “blue”, fill = “lightblue”) + facet_wrap(~reorder(CONTINENT, -TFR, median), nrow = 6, strip.position = “right”) + theme(strip.placement = “outside”, strip.background = element_blank(), strip.text = element_text(face = “bold”)) ### Density curve geom_density(color = “red”) ### Violin plot geom_violin(adjust = 6) # change bandwidth ### Ridgeline plot gr <- ggplot(world, aes(x = GDP, y = reorder(CONTINENT, -GDP, median))) + geom_density_ridges(fill = “blue”, alpha = .5)

39.3 Chapter 6 Rounding noraml distrubution

39.3.1 Q-Q plot 

3.1.1 What is a Q-Q plot?
If the line is not 45 degrees, then it is not normally distrubtted

3.1.2 How to draw
qqnorm(x)

3.1.3 DIY Q-Q plot
 qx <- quantile(x) 
 qn <- quantile(n)
 plot(qn, qx, pch = 16)
 mod <- lm(c(qx[2], qx[4])~c(qn[2], qn[4]))
 abline(mod, col = "red")

39.3.2 Density Curve + Noraml Curve 

3.2.1 How to draw
      library(tidyverse)
      df <- data.frame(x = rnorm(1000, 50, 10))
      ggplot(df, aes(x)) + geom_histogram(aes(y = ..density..),
                                          fill = "lightblue",
                                          color = "black") +
        geom_density(lwd = 1.5) +
        stat_function(fun = dnorm, args = list(mean = 50, sd = 10), color = "red", lwd = 1.5)

39.3.3 Shapiro Wilk test 

3.3.1 What is this
Null hypothesis: data is normally distributed
Alternative hypothesis: data is not normally distributed

3.3.2 How 
> shapiro.test(x)
if P-value is very small, then the null hypothesis could be rejected.

39.4 Chapter 7 Grammar of Graphics

39.4.1 Layers 

    g <- ggplot() + geom_point(data = df1, aes(x,y)) +
             geom_col(data = df2, aes(num, height),
             fill = "green") + 
             geom_boxplot(data = df3, aes(1, score)) +
             geom_line(data = df4, aes(time, dist),
             color = "red")

39.4.2 Coord 

g+coord_polar()

39.4.3 Facet 

g + facet_grid(num~height)

39.4.4 Theme 

g+theme_wsj()

39.4.5 Mapping 

aes(x = x, y = y, color = "color", fill = "fill")

39.4.7 Code style 

    ggplot (data =  <DATA> ) +  
        <GEOM_FUNCTION> (mapping = aes( <MAPPINGS> ),  
        stat = <STAT> , position = <POSITION> ) +      
        <COORDINATE_FUNCTION> + 
        <FACET_FUNCTION>  + 
        <SCALE_FUNCTION>  + 
        <THEME_FUNCTION>

39.5 Chapter 8 Categorical Variables

39.5.1 Types of data 

    nominal – no fixed category order
    ordinal – fixed category order
    (“real”) discrete, small # of possibilities
    Not always clearcut: nominal vs. ordinal, ordinal vs. discrete, and...
    Sometimes numbers = nominal, not discrete
    
    5.1.1 Ordinal data
          Sort in logical order of the categories (left to right, starting at bottom or top, etc.)
    5.1.2 Nominal data
          Sort from highest to lowest count (left to right, or top to bottom, etc.)
    
    5.1.3 Discrete data
          * Cleveland dot plot:
            ggplot(africa, aes(x = GDP, y = fct_reorder(COUNTRY, GDP))) +
              geom_point(color = "blue")
            hint: Set x and y axis different: scales = "free_y", space = "free_y"

39.5.2 Plotting Categorical data 

    5.2.1 Recoding factor levels: 
          
          1) setting levels by hand:
          levels(x) <- c("Physics", "Math", "Chemistry")
          
          2) fct_recode: change names of levels
          x <- factor(c("G234", "G452", "G136"))  
          y <- fct_recode(x, Physics = "G234", Math = "G452", Chemistry = "G136")  
          
          3) fct_inorder: set level order of x to row order
          ggplot(df, aes(x = fct_inorder(temperature), y = count))
          
          4) fct_relevel: manually set the order of levels of x
          x <- c("A", "B", "C", "move1", "D", "E", "move2", "F")  
          fct_relevel(x, "move1", "move2")
          
          5) fct_reorder: reorder x by y
          ggplot(pack1, aes(fct_reorder(color, count), count)) +   
          geom_col() 
          
          6) fct_infreq: order the levels of x by decreasing frequency
          ggplot(df, aes(fct_infreq(mmcolor))) +   
          geom_bar() 
          
          7) fct_rev: reverse the order of factor levels of x
          ggplot(df, aes(fct_rev(fct_infreq(mmcolor)))) +   
          geom_bar()
          
          8) fct_explicit_na: turn NAs into a real factor level
          
    5.2.2 Binning and rebinning data
          df %>%   
            group_by(Class) %>%   
            summarize(Freq = sum(Freq)) %>%   
            ggplot(aes(Class, Freq)) +   
            geom_col(color = "grey50", fill = "lightblue") +  
            theme_grey(16)
          
    5.2.3 Percentages
          df %>%   
            gro up_by(Class) %>%   
            summarize(Freq = sum(Freq)) %>%   
            mutate(prop = Freq/sum(Freq))

39.6 Chapter 10 Tidying data

Tidy: 1 variable per column, 1 observation per row

39.7 Chapter 11 Dependency relationships

39.7.1 Interactive plot 

    1) ggplot
    library(plotly)
    ggplotly(g)
    
    2) base r
    plot_ly(world, x = ~GDP, y = ~TFR, color = ~CONTINENT, text = ~COUNTRY, hoverinfo = 'text')

39.7.2 Dealing with overplotting 

    7.2.1 Open circles:
    geom_point(pch = 21)
    
    7.2.2 Alpha blending: 
    geom_point(alpha = .05, stroke = 0) 
    
    7.2.3 Smaller dots:
    size = .05
    shape = "."
    
    7.2.4 Subset:
    ggplot(binnedmovies, aes(votes, rating)) + 
    geom_point(alpha = .1) +
      facet_wrap(~mybin, scales = "free_x")
      
    7.2.5 10% highest number:
    bin10 <- binnedmovies %>% filter(mybin == 10)
    ggplot(bin10, aes(votes, rating)) + geom_point()

39.7.3 Heatmap 

    7.3.1 Parameters:
    bindwidth = c(10, 10) changes large or small of the shape
    alpha - 0.4 changes color dark or light
    size = 2 changes large or small of the dots
    scale_fill_gradient(low = "grey", high = "purple") changes color
   
    7.3.2 Square heatmap:
    g1 + geom_bin2d(binwidth = c(5, 5)) # no dots, dark color
    g1 + geom_bin2d(binwidth = c(5, 5), alpha = .4) + geom_point(size = 2)
    
    7.3.3 Hex heatmap:
    g1 + geom_hex(binwidth = c(5, 5))

39.7.4 Density estimate contour 

     g +  geom_density_2d(bins = 5)
     
     contour lines: 
     contour(f1, add = T)
     g + geom_contour(aes(z = z))

39.8 Chapter 12 Continuous + Categorical

39.8.1 Graphiical data analysis with R 

   Examining Continuous Variables
   Displaying Categorical Data
   Looking for Structure: Dependency Relationships and Associations
   Combining Continuous and Categorical Data
   Investingating Multivariate Continuous Data
   Studying Multivariate Categorical Data

39.8.2 Factes

facet_wrap(~categorical variable, scales = “free_x”) “free_x” means x axises are not the same

39.8.3 Mapping options

Continuous: x-axis, y-axis, color (not so great), size (not so great) Categorical: color, facets (rows, columns), shape (maybe)

39.9 Parallel Coordinates (Multivariate continuous data)

9.1 3D scatterplot 3D scatterplot: library(scatterplot3d)
scatterplot3d(df\(x, df\)y, df$z, pch = 16, color = “blue”)

interactive 3D: library(plotly)
plot_ly(df, x = ~x, y = ~y, z = ~z, mode = “markers”,
marker = list(size = 4)) %>% add_markers()

39.9.1 Slope graph: ggplot2::geom_line()

ggplot(tidydf, aes(x = var, y = value, group = ID)) +
geom_line()

   9.2.1 Standardize data
         mutate(value = standardize(value)) %>% ungroup()
   9.2.2 Rescale data to [0,1]
         mutate(value = scales::rescale(value)) 
   9.2.3 Parameters: GGally::ggparcoord() (static, ggplot2)
         1. Scale: ggparcoord(mystates, columns = 2:9, scale = " ")
            1) scale = "globalminmax"
            2) scale = “std” (default)
            3) scale = "uniminmax"
         
         2. Reorder:
            ggparcoord(mystates, columns = c(2, 4, 6, 8, 3, 5, 7, 9))
         
         3. alpha: changes color light or dark
         
         4. vline: add vertical lines
            geom_vline(xintercept = 1:4, color = "lightblue")
            
         5. splines: make lines softer
            gparcoord(df, columns = 1:4, scale = "globalminmax",  
            splineFactor = 10) 
         
         6. group: add color to different groups
            ggparcoord(mystates, columns = 2:9, groupColumn = 10)
         
         7. parcoords:
         MASS:: parcoord() (static, base)
         parcoords::parcoords() (interactive)  
         
            # See: http://www.buildingwidgets.com/blog/2015/1/30/week-04-interactive-parallel-coordinates-1 
            # devtools::install_github("timelyportfolio/parcoords")  
            library(parcoords)  
            mystates  %>% arrange(Region) %>%  
              parcoords(  
                rownames = F   
                , brushMode = "1D-axes"  
                , reorderable = T  
                , queue = T  
                , alpha = .5
                )  
                
         8. color:
             parcoords(mystates  
                , rownames = F   
                , brushMode = "1D-axes"  
                , reorderable = T  
                , queue = T  
                , color = list(  
                  colorBy = "Region"  
                  ,colorScale = "scaleOrdinal"  
                  ,colorScheme = "schemeCategory10"  
                  )  
                , withD3 = TRUE  
                )    
            
         9. highlighting a tred: Murder>11 is highlighted here:
         mystates <- mystates %>%  
             mutate(color = factor(ifelse(Murder > 11, 1, 0))) %>%   
             arrange(color)  
         ggparcoord(mystates, columns = 2:9, groupColumn = "color") + 
             scale_color_manual(values = c("grey70", "red")) +
             coord_flip() + guides(color = FALSE) +  
             ggtitle("States with Murder Rate > 11 (per 100000) in red")

39.10 Biplots

39.10.1 Principal components analysis 

     pca <- prcomp(ratings[,2:7], scale. = TRUE)

39.10.2 Biplot 

     1) ggplot
     ggplot(df_pca, aes(PC1, PC2, label = country)) +
     geom_point() +
     geom_text(nudge_y = .2) +
     coord_fixed()
     
     2) base r
     draw_biplot(ratings, "living_standard")

39.10.3 Scatterplot of scaled data

39.11 Multivariate Categorical data

39.11.1 Stacked bar chart 

ggplot(cases, aes(x = Age, fill = Favorite)) + 
  geom_bar() + 
  scale_fill_manual(values = c("#ff99ff", "#cc9966"))

39.11.2 Group bar chart 

ggplot(cases, aes(x = Age, fill = Favorite)) +
  geom_bar(position = "dodge") +

39.11.3 Cleveland dot plot 

1) Setting theme
   theme_dotplot <-
     theme_bw(16) +
     theme(axis.ticks.y = element_blank(), panel.grid.major.x = element_blank(),
           panel.grid.major.y = element_line(size = 0.5), panel.grid.minor.x = element_blank())
2) geom_point

39.11.4 Mosaic plot 

1) two variables
   library(vcd)
   tidyexp$Group <- fct_rev(tidyexp$Group)
   mosaic(Group ~ Age, direction = c("v", "h"), tidyexp,
          highlighting_fill = c("grey80", "cornflowerblue"))
2) three variables
   vcd::mosaic(Favorite ~ Music + Age, counts3,
   direction = c("v", "v", "h"), # <- order: Music ("v"), Age ("v"), Favorite ("h")
   highlighting_fill = icecreamcolors)
   
3) Mosaic pairs plot
   pairs(table(cases[,2:4]), highlighting = 2)
   
4) Mosaic mobility plot
   UKmob$mob <- factor(UKmob$mob, levels = 4:-4)
   UKmob$Father <- factor(UKmob$Father, levels = levels(Yamaguchi87$Father))
   vcd::mosaic(mob ~ Father, UKmob, 
               direction = c("v", "h"),
               tl_labels = c(FALSE, TRUE),  # move labels to bottom  
               rot_labels = c(0,0,0,0),   # all horizontal
               highlighting_fill = fills9,
               main = "Downward mobility")
 5) Change of names:
    library(vcd)
    library(vcdExtra)
    library(tidyverse)
    
    foodorder <- Alligator %>% group_by(food) %>% summarize(Freq = sum(count)) %>% 
      arrange(Freq) %>% pull(food)
    
    ally <- Alligator %>% 
      rename(Freq = count) %>% 
      mutate(size = fct_relevel(size, "small"),
             food = factor(food, levels = foodorder),
             food = fct_relevel(food, "other"))
    
    vcd::mosaic(food ~ sex + size, ally,
           direction = c("v", "v", "h"),
           highlighting_fill= RColorBrewer::brewer.pal(5, "Accent"))
  
  6) Treemap
     library(treemap)
     # example from ?treemap
     data(GNI2014)
     treemap::treemap(GNI2014,
            index=c("continent", "iso3"),
            vSize="population",
            vColor="GNI",
            type="value",
            format.legend = list(scientific = FALSE, big.mark = " "))
   
   7) Spine plot:  mosaic plot with straight, parallel cuts in one dimension (“spines”) and only one variable cutting in the other direction
      vcd::mosaic(food ~ sex + size, ally,
      direction = c("v", "v", "h"),
      highlighting_fill= RColorBrewer::brewer.pal(5, "Accent"))
   
   8) ggplot
      fillcolors <- brewer.pal(3, "Blues")
      h <- housing %>% group_by(Type, Infl, Cont) %>% 
        mutate(RelFreq = Freq/sum(Freq)) %>% ungroup()
      ggplot(h, aes(x = Infl, y = RelFreq, fill = Sat)) +
        geom_col() +
        facet_grid(Cont ~ Type) +
        scale_fill_manual(values = fillcolors) +
     theme_classic()

39.11.5 Chi-square test of independence 

   x<- chisq.test(matrix, correct = F)
   x$observed
   x$expected
   x

39.12 Alluvial form 

1) Structure: stratum, load, alluvium
2) Basic code:
   library(ggalluvial)
   ggplot(df, aes(axis1 = Class1, axis2 = Class2, y = Freq)) +
     geom_alluvium(color = "black") +
     geom_stratum() +
     geom_text(stat = "stratum", aes(label = paste(after_stat(stratum), "\n", after_stat(count)))) +
     scale_x_discrete(limits = c("Class1", "Class2"))
3) Separate code:
   3.1) Strata only:
        ggplot(df, aes(axis1 = Class1, axis2 = Class2, y = Freq)) +
        #  geom_alluvium(color = "black") +
          geom_stratum() +
          geom_text(stat = "stratum", aes(label = paste(after_stat(stratum), "\n", after_stat(count)))) +
          scale_x_discrete(limits = c("Class1", "Class2"))
   3.2) Alluvia only:
         ggplot(df, aes(axis1 = Class1, axis2 = Class2, y = Freq)) +
           geom_alluvium(color = "black") +
         #  geom_stratum() +
         #  geom_text(stat = "stratum", aes(label = paste(after_stat(stratum), "\n", after_stat(count)))) +
           scale_x_discrete(limits = c("Class1", "Class2"))
4) geom_flow:
    df2 <- data.frame(Class1 = c("Stats", "Math", "Stats", "Math", "Stats", "Math", "Stats", "Math"),
              Class2 = c("French", "French", "Art", "Art", "French", "French", "Art", "Art"),
              Class3 = c("Gym", "Gym", "Gym", "Gym", "Lunch", "Lunch", "Lunch", "Lunch"),
              Freq = c(20, 3, 40, 5, 10, 2, 5, 15))
    ggplot(df2, aes(axis1 = Class1, axis2 = Class2, axis3 = Class3, y = Freq)) +
      geom_flow(color = "black") +
      geom_stratum() +
      geom_text(stat = "stratum", aes(label = paste(after_stat(stratum), "\n", after_stat(count)))) +
      scale_x_discrete(limits = c("Class1", "Class2", "Class3"))
5) color by first variable:
    library(vcdExtra)
    ggplot(Yamaguchi87, aes(y = Freq, axis1 = Father, axis2 = Son)) +
      geom_flow(aes(fill = Father), width = 1/12) +
      geom_stratum(width = 1/12, fill = "grey80", color = "black") +
      geom_label(stat = "stratum", aes(label = after_stat(stratum))) +
      scale_x_discrete(limit = c("Father", "Son"), expand = c(.05, .05)) +
      scale_y_continuous(expand = c(.01, 0)) +
      guides(fill = FALSE) +
      theme_classic()

39.13 Color

  1. Continuous data:
    +scale_color_viridis_c() or fill

    recolorbrewer:

    • scale_color_distiller(palette = “PuBu”)

    reverse palette order with direction = 1

  2. Continuous data: Create your own sequential: +scale_color_gradient(low = “white”, high = “red”) Create your own diverging: +scale_color_gradient2(low = “blue”, mid = “white”, high = “red”)

  3. Discrete data: +scale_color_viridis_d()

    recolor brewer: +scale_color_brewer(palette = “PuBu”)

    Create your own: +scale_color_manual(values = c(“red”, “yellow”, “blue”)

4) Discrete ordinal data:
   library(RColorBrewer)
   colors<- brewer.pal(4, "Reds)
   barplot(1:4, col = colors)

39.14 Time Series

39.14.1 Multiple time series: line 

    df <- df %>% gather(key = TYPE, value = RATE, -DATE) %>%
      mutate(TYPE = forcats::fct_reorder2(TYPE, DATE, RATE))# puts legend incorrect order

39.14.2 filling: 

    ggplot(df, aes(DATE, RATE, fill = TYPE))

39.14.3 plot: 

    plot()
    monthplot()

39.14.4 label by day of week 

      ggplot(christmas, aes(Date, Gross)) +
       geom_label(aes(label = wday(Date, label = TRUE)))
    label by day of month 
      geom_label(data = christmas, 
       aes(x = Date, y = Gross/1000000 + .06, 
           label = day(Date)))

39.14.5 Highlight the abnormality 

      g + annotate("rect", xmin = start, xmax = end,
               ymin = -Inf, ymax = Inf, fill = "green",
               alpha = .2) +
      annotate("text", x = end + 2,
               y = 1500000, label = "Dec 24 - Jan 2",
               color = "green", hjust = 0)

39.14.6 Scale the data 

      mutate(index = round(100*value/value[1], 2))

39.14.7 Gaps 

      mydat$Date <- as.Date(paste0(mydat$Year, "-", as.character(mydat$Q*3), "-30"))
      scale_x_date(limits = c(as.Date("2012-02-01"), as.Date("2014-12-31")), date_breaks = "6 months", date_labels = "%b %Y")

39.14.8 Leave gaps 

       geom_path(aes(group = `Order method type`)) +
       scale_x_date(limits = c(as.Date("2012-02-01"), as.Date("2014-12-31")),
           date_breaks = "6 months", date_labels = "%b %Y")

39.15 Dates

39.15.1 Convert to date class 

     1) Convert
        date<- as.Date(character)
        # class(date) >> Date
     2) Specifying the format
        as.Date("Thursday, January 6, 2005", format = "%A, %B %d, %Y")
        ## "2005-01-06"
     3) Parse_date
        as.Date("1/12/2019", format="%m/%d/%y")
        ## "2020-01-12"
        readr::parse_date("1/12/2019", format="%m/%d/%y")
        ## NA
       
        as.Date("Thursday, January 6, 2005", format = "%A, %B %d, %Y")
        ## "2005-01-06"
        readr::parse_date("Thursday, January 6, 2005", format = "%A, %B %d, %Y")
        ## Error: Invalid %A auto parser

39.15.2 Lubridate: ymd(), ydm(), mdy(), myd(), dmy(), dym() 

     1) mdy
        lubridate::mdy("April 13, 1907")
        ## "1907-04-13"
     2) as.Date
        as.Date("2017-11-02") - as.Date("2017-01-01")
        ## Time difference of 305 days
     3) Sys.Date()
        ## "2021-10-25"
        class(Sys.Date())
        ## "Date"
     4) today <- Sys.Date()
        lubridate::year(today)
        ## 2021
        
        lubridate::yday(today)
        ## 298
        
        lubridate::month(today, label = TRUE)
        ## Oct
        ## 12 Levels: Jan < Feb < Mar < Apr < May < Jun < Jul < Aug < Sep < ... < Dec

39.15.3 Base R functions 

     1) today <- Sys.Date()
        weekdays(today, abbreviate = TRUE)
        ## "Mon
     2) months(today, abbreviate = TRUE)
        ## "Oct"
     3) quarters(today)
        ## "Q4"

39.16 Missing data

  1. Missing values by column colSums(is.na(mycars)) %>% sort(decreasing = TRUE)

  2. Missing values by row rowSums(is.na(mycars)) %>% sort(decreasing = TRUE)

  3. Plotting missing data ggplot(tidycars, aes(x = key, y = fct_rev(id), fill = missing))

  4. Heatmap missing_data.frame(mycars)

  5. Missing values by variable ggplot(tidycars, aes(x = key, y = fct_rev(id), fill = value)) + geom_tile(color = “white”) + scale_fill_gradient(low = “grey80”, high = “red”, na.value = “black”)

  6. Missing values by variable mi::missing_data.frame(mycars)

  7. Aggregate missing patterns plot_missing(mycars, percent = FALSE)

  8. Number missing by day of month · add day of week info missing <- missing %>% mutate(dayofweek = weekdays(as.Date(key), abbreviate = FALSE)) · correct day order daysinorder <- c(“Monday”, “Tuesday”, “Wednesday”, “Thursday”, “Friday”, “Saturday”, “Sunday”)

    · reorder dayofweek missing\(dayofweek <- factor(missing\)dayofweek, levels = daysinorder) · choose colors daycolors <- c(rep(“#cbc9e2”, 5), rep(“#2b8cbe”, 2))

    · plot missing values by day, weekday/weekend colors ggplot(missing, aes(x = key, y = sum.na, fill = dayofweek)) + geom_col() + ggtitle(“Number of missing values by day”) + scale_fill_manual(values = daycolors) + xlab(““) + ylab(”Number of missing station values (out of 349)“) + theme_classic() + theme(axis.text.x = element_text(angle = 90))