Cheat Sheet

Hiding Render Output

• Hide source code: {r, echo=FALSE}
• Hide text output: {r, results = FALSE} #Summary Tables
• Hide messages: {r, message=FALSE} #Package Loading
• Hide warning messages: {r, warning=FALSE}
• Hide Plots: {r, fig.show = hide}

Data Cleaning

• select() :which extracts columns from a data frame

• filter() :which extracts rows from a data frame

  • Filter NAs: filter(!is.na(x))
  • filter(hair == “black” & eye == “blue”) would dive us columns that are both black hair and blue eyes
  • filter(hair == “black” | eye = “blue”) would give us columns where hair is black or eyes are blue

• arrange() :which moves important rows to the top of a data frame

  • Will naturally put in ascending/alphabetical order.

  • arrange(desc(x)) will put dataframe in descending order

Dataset Wrangling

• Combine column from one dataset to another: - I have column tree in ds1 that I want to add to ds2

  • cbind command:
    cbind(ds2, x = ds1$tree)

  • _join commands

• Strsplit: Split up one name into multiple:

  • Example: BO_19_2938 → BO, 19, 2938 (so 3 columns)
    data %>% strsplit(split = "_")

• Substring: Just take a portion of the name

  • Example: BO_19_1000
    Data %>% substr(1,5) —>Would give us BO_19

• Filter columns containing “x”

  data %>% filter(grepl('x', column))

• Rename Column:
  Data %>% rename("New" = "Old")

• Rename Variables In a Column”
  mutate(column = recode(column, "Old" = "Blind New"))

• If value is NA, make it X (can be numeric or “character”)

  mutate(col= ifelse(is.na(col), "X", col))

• Make new column based on old column:
  mutate(new_col = old_col*6) —>If old column had “3”, new column would have “18”

• Average values based on column (Ex. Triplicates in qPCR)“
  group_by(non-numeric things) %>%
summarize_if(is.numeric, mean) %>%
ungroup()

• Add two datasets with same column headings vertically:
  rbind(d1,d2)

• Create new column using mutate() and case_when:
  Df %>% mutate(new_column = case_when(other_column == "A"~ "What_you_want_in_new_col")

• Create new column using mutate, case_when, and if_else

  Df %>% mutate(new_column = case_when(other_column == "A"~ "What_you_want_in_new_col", TRUE ~ else-output-value)

GGPLOT

• Tilting x axis labels:
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

• Centering Title:
  theme(plot.title = element_text(hjust = 0.5))

• Remove legend:
  theme(legend.position = "none")

• Add more ticks to x or y axis:
  scale_x_continuous(n.breaks=10) // scale_y_continuous

• Stack Plots with ggarrange():
  ggarrange(fig1, fig2, fig3 + font("x.text", size = 10), ncol = 1, nrow = 3)

  • Final plot is 1 column and 2 rows

• Edit axis or title labels aesthetically:
  theme(axis.title.y = element_text(hjust = 1, color = "red, face = "bold)

• Set colors by factor:
  scale_color_manual(values = c("setosa" = "purple","versicolor" = "red"))

• Set shapes by factor:
  scale_shape_manual(values = c("setosa" = 1, "versicolor" = 2))

• Add annotation:
  annotate("text", x = 100, y =200, label = "WORD", color = "blue", fontface=2, size = 6.4)

• Add Arrows:
  annotate( geom = "curve", x = 100, y = 200, xend = 200, yend = 300, curvature = .45, arrow = arrow(length = unit(2, "mm")))

Statistics

• Linear model: Relationship between infant_mortality (x) and life_expectancy (y) from gapminder dataset:
  lm(infant_mortality\~life_expectancy, data = gapminder)

• Add R2 to graph

  bo2 %>% ggplot(aes(x = lag, y = copies_mL)) + geom_point() + stat_smooth(method = "lm")+ stat_correlation(method = "pearson")+ labs(title = "LM DustxCopies (BO+1)", x = "log(dust)")

Exporting

• Save as RDS file in specific folder:
  saveRDS(data, file = "Desktop/cheese/newdata.rds") would save the file newdata.rds in a folder called cheese on the desktop
  readRDS("Desktop/cheese/newdata.rds") would load it back in

• Save plot as png:
  png(file = "file/location/plot.png")
Gpplot code
dev.off()

Random

• Make a nice table:
  table<- data.frame(Height, Width)
kable(table)