Sure we all love ggplot2. It allows to create visualizations by simply conceptualizing them — as long as you learn the syntax.

But what about tables?

R packages gt and gtExtras have you covered.

Let’s turn a simple data.frame into a gt table. This will require familiarity with the Tidyverse.

library(gt)
library(tidyverse)
data <- data.frame(
    text = c("Eufemia", "Antonia", "Michele", "Gerardo"),
    year = c("1955", "1967", "1968", "1972")
    ) %>% gt()

data

text	year
Eufemia	1955
Antonia	1967
Michele	1968
Gerardo	1972

Understanding the gt philosophy

The gt package follows a grammar of tables approach, similar to how ggplot2 implements the grammar of graphics. Every table has distinct components that can be styled and modified independently.

Table anatomy

A gt table consists of several parts:

Table header: Title and subtitle
Column labels: Column headers and spanners
Table body: The actual data
Summary rows: Aggregated data
Table footer: Source notes and footnotes

# Let's build a more complete example
clinical_data <- tribble(
  ~Treatment, ~N, ~Response_Rate, ~CI_Lower, ~CI_Upper, ~p_value,
  "Drug A",   125, 0.72,          0.64,      0.80,      0.001,
  "Drug B",   130, 0.65,          0.57,      0.73,      0.023,
  "Placebo",  128, 0.45,          0.37,      0.53,      NA
)

clinical_data %>%
  gt() %>%
  tab_header(
    title = "Clinical Trial Results",
    subtitle = "Phase III Randomized Controlled Trial"
  )

Clinical Trial Results
Phase III Randomized Controlled Trial
Treatment	N	Response_Rate	CI_Lower	CI_Upper	p_value
Drug A	125	0.72	0.64	0.80	0.001
Drug B	130	0.65	0.57	0.73	0.023
Placebo	128	0.45	0.37	0.53	NA

Essential formatting

Column formatting

The power of gt lies in its formatting functions. Let’s properly format our clinical data:

clinical_table <- clinical_data %>%
  gt() %>%
  tab_header(
    title = "Clinical Trial Results",
    subtitle = "Phase III Randomized Controlled Trial"
  ) %>%
  fmt_percent(
    columns = c(Response_Rate, CI_Lower, CI_Upper),
    decimals = 1
  ) %>%
  fmt_number(
    columns = p_value,
    decimals = 3
  ) %>%
  sub_missing(
    columns = everything(),
    missing_text = "—"
  )

clinical_table

Clinical Trial Results
Phase III Randomized Controlled Trial
Treatment	N	Response_Rate	CI_Lower	CI_Upper	p_value
Drug A	125	72.0%	64.0%	80.0%	0.001
Drug B	130	65.0%	57.0%	73.0%	0.023
Placebo	128	45.0%	37.0%	53.0%	—

Column labels

Professional tables need clear, formatted column labels:

clinical_table <- clinical_table %>%
  cols_label(
    Treatment = "Treatment Group",
    N = "Sample Size",
    Response_Rate = "Response Rate",
    CI_Lower = "95% CI",
    CI_Upper = "",  # Empty label for upper CI bound
    p_value = "p-value"
  ) %>%
  cols_merge(
    columns = c(CI_Lower, CI_Upper),
    pattern = "{1}–{2}"
  )

clinical_table

Clinical Trial Results
Phase III Randomized Controlled Trial
Treatment Group	Sample Size	Response Rate	95% CI	p-value
Drug A	125	72.0%	64.0%–80.0%	0.001
Drug B	130	65.0%	57.0%–73.0%	0.023
Placebo	128	45.0%	37.0%–53.0%	—

Advanced styling

Conditional formatting

Highlight significant results and add visual hierarchy:

clinical_table <- clinical_table %>%
  tab_style(
    style = list(
      cell_text(weight = "bold"),
      cell_fill(color = "#E8F4F8")
    ),
    locations = cells_body(
      columns = everything(),
      rows = p_value < 0.05
    )
  ) %>%
  tab_style(
    style = cell_text(weight = "bold"),
    locations = cells_column_labels(everything())
  )

clinical_table

Clinical Trial Results
Phase III Randomized Controlled Trial
Treatment Group	Sample Size	Response Rate	95% CI	p-value
Drug A	125	72.0%	64.0%–80.0%	0.001
Drug B	130	65.0%	57.0%–73.0%	0.023
Placebo	128	45.0%	37.0%–53.0%	—

Adding summary rows

Include summary statistics directly in your table:

# Create a grouped dataset for summary rows
grouped_data <- tibble(
  Group = c("Active", "Active", "Control"),
  Treatment = c("Drug A", "Drug B", "Placebo"),
  N = c(125, 130, 128),
  Response_Rate = c(0.72, 0.65, 0.45)
)

summary_table <- grouped_data %>%
  gt(groupname_col = "Group") %>%
  summary_rows(
    groups = "Active",
    columns = N,
    fns = list(
      Total = ~sum(., na.rm = TRUE)
    ),
    formatter = fmt_number,
    decimals = 0
  ) %>%
  summary_rows(
    groups = "Active",
    columns = Response_Rate,
    fns = list(
      Mean = ~mean(., na.rm = TRUE)
    ),
    formatter = fmt_percent,
    decimals = 1
  ) %>%
  tab_header(
    title = "Treatment Group Summary"
  )

summary_table

Treatment Group Summary
	Treatment	N	Response_Rate
Active
	Drug A	125	0.72
	Drug B	130	0.65
Total	—	255	—
Mean	—	—	68.5%
Control
	Placebo	128	0.45

Real-world example: Systematic review table

Let’s create a publication-ready table for a systematic review:

# Simulate systematic review data
systematic_review <- tribble(
  ~Study, ~Year, ~Design, ~N, ~Intervention, ~Outcome, ~Effect_Size, ~CI_95, ~Quality,
  "Smith et al.", 2020, "RCT", 250, "Drug A", "Mortality", 0.85, "0.72–0.99", "High",
  "Jones et al.", 2021, "RCT", 180, "Drug A", "Mortality", 0.78, "0.65–0.93", "High",
  "Brown et al.", 2019, "Cohort", 500, "Drug A", "Mortality", 0.91, "0.83–1.01", "Moderate",
  "Davis et al.", 2022, "RCT", 320, "Drug A", "Mortality", 0.82, "0.71–0.95", "High",
  "Wilson et al.", 2021, "Cohort", 450, "Drug A", "Mortality", 0.88, "0.77–1.02", "Moderate"
)

systematic_review %>%
  gt() %>%
  tab_header(
    title = md("**Systematic Review of Drug A for Mortality Reduction**"),
    subtitle = "Studies published 2019–2022"
  ) %>%
  cols_label(
    Study = "Study",
    Year = "Year",
    Design = "Study Design",
    N = "Sample Size",
    Intervention = "Intervention",
    Outcome = "Primary Outcome",
    Effect_Size = "Hazard Ratio",
    CI_95 = "95% CI",
    Quality = "Quality Assessment"
  ) %>%
  tab_spanner(
    label = "Study Characteristics",
    columns = c(Study, Year, Design, N)
  ) %>%
  tab_spanner(
    label = "Results",
    columns = c(Intervention, Outcome, Effect_Size, CI_95)
  ) %>%
  fmt_number(
    columns = N,
    sep_mark = ",",
    decimals = 0
  ) %>%
  tab_style(
    style = cell_fill(color = "#F0F0F0"),
    locations = cells_body(rows = Design == "Cohort")
  ) %>%
  tab_footnote(
    footnote = "RCT = Randomized Controlled Trial",
    locations = cells_column_labels(columns = Design)
  ) %>%
  tab_source_note(
    source_note = "Quality assessment based on Cochrane Risk of Bias tool"
  )

Study Characteristics				Results				Quality Assessment
Systematic Review of Drug A for Mortality Reduction
Studies published 2019–2022
Study	Year	Study Design¹	Sample Size	Intervention	Primary Outcome	Hazard Ratio	95% CI	Quality Assessment
Smith et al.	2020	RCT	250	Drug A	Mortality	0.85	0.72–0.99	High
Jones et al.	2021	RCT	180	Drug A	Mortality	0.78	0.65–0.93	High
Brown et al.	2019	Cohort	500	Drug A	Mortality	0.91	0.83–1.01	Moderate
Davis et al.	2022	RCT	320	Drug A	Mortality	0.82	0.71–0.95	High
Wilson et al.	2021	Cohort	450	Drug A	Mortality	0.88	0.77–1.02	Moderate
Quality assessment based on Cochrane Risk of Bias tool
¹ RCT = Randomized Controlled Trial

Integration with gtExtras

The gtExtras package extends gt with additional functionality:

library(gtExtras)

# Create a table with sparklines
trend_data <- tibble(
  Metric = c("Enrollment", "Retention", "Compliance"),
  Current = c(125, 0.85, 0.92),
  Trend = list(
    c(100, 105, 110, 115, 120, 125),
    c(0.80, 0.82, 0.83, 0.84, 0.84, 0.85),
    c(0.90, 0.91, 0.91, 0.92, 0.92, 0.92)
  ),
  Change = c("+25%", "+5%", "+2%")
)

trend_table <- trend_data %>%
  gt() %>%
  # Format Current column based on row content
  text_transform(
    locations = cells_body(columns = Current),
    fn = function(x) {
      ifelse(
        trend_data$Metric %in% c("Retention", "Compliance"),
        paste0(as.numeric(x) * 100, "%"),
        x
      )
    }
  ) %>%
  gt_plt_sparkline(
    column = Trend
  ) %>%
  tab_header(
    title = "Clinical Trial Metrics Dashboard",
    subtitle = "6-Month Performance Summary"
  ) %>%
  tab_style(
    style = cell_text(color = "#28A745", weight = "bold"),
    locations = cells_body(
      columns = Change
    )
  )

trend_table

Clinical Trial Metrics Dashboard
6-Month Performance Summary
Metric	Current	Trend	Change
Enrollment	125.00		+25%
Retention	85%		+5%
Compliance	92%		+2%

Alternatively, create a simpler example with bar charts:

# Example with gt_plt_bar
performance_data <- tibble(
  Department = c("Cardiology", "Oncology", "Neurology", "Pediatrics"),
  Enrollment_Rate = c(85, 92, 78, 88),
  Target = rep(80, 4)
)

performance_data %>%
  gt() %>%
  gt_plt_bar(
    column = Enrollment_Rate,
    color = "#1E88E5",
    width = 50
  ) %>%
  fmt_number(
    columns = c(Enrollment_Rate, Target),
    decimals = 0,
    suffix = "%"
  ) %>%
  tab_header(
    title = "Department Enrollment Performance",
    subtitle = "Q4 2023 Results vs. Target"
  ) %>%
  tab_style(
    style = cell_fill(color = "#E8F4F8"),
    locations = cells_body(
      columns = everything(),
      rows = Enrollment_Rate >= Target
    )
  )

Department Enrollment Performance
Q4 2023 Results vs. Target
Department	Enrollment_Rate	Target
Cardiology		80
Oncology		80
Neurology		80
Pediatrics		80

Export options

Save as image

For manuscripts and presentations:

# First install webshot2 if needed
# install.packages("webshot2")

# Save as PNG
clinical_table %>%
  gtsave("clinical_results_table.png", expand = 10)

# Save as PDF
clinical_table %>%
  gtsave("clinical_results_table.pdf")

Note

Saving tables as images requires the webshot2 package. Install it with install.packages("webshot2") if you haven’t already.

HTML output

For web publishing and interactive documents:

# Save as standalone HTML
clinical_table %>%
  gtsave("clinical_results_table.html")

# Or get the HTML code
html_code <- clinical_table %>%
  as_raw_html()

Word and LaTeX

For journal submissions:

# Export to RTF (which can be opened in Word)
clinical_table %>%
  gtsave("clinical_results.rtf")

# For LaTeX
latex_code <- clinical_table %>%
  as_latex()

# Save to file
writeLines(latex_code, "clinical_results.tex")

Tip

For Word documents, the RTF format provides the best compatibility. Most word processors can open RTF files while preserving table formatting.

Best practices

Design principles

Less is more: Remove unnecessary gridlines and borders
Hierarchy matters: Use visual weight to guide the reader
Consistency: Maintain uniform formatting throughout

# Clean, publication-ready style
clean_table <- clinical_data %>%
  gt() %>%
  tab_header(
    title = "Clinical Trial Results"
  ) %>%
  fmt_percent(columns = c(Response_Rate, CI_Lower, CI_Upper)) %>%
  fmt_number(columns = p_value, decimals = 3) %>%
  cols_merge(
    columns = c(CI_Lower, CI_Upper),
    pattern = "{1}–{2}"
  ) %>%
  tab_options(
    table.border.top.width = px(0),
    table.border.bottom.width = px(2),
    table.border.bottom.color = "black",
    column_labels.border.top.width = px(2),
    column_labels.border.top.color = "black",
    column_labels.border.bottom.width = px(2),
    column_labels.border.bottom.color = "black",
    data_row.padding = px(6),
    heading.align = "left"
  ) %>%
  opt_table_lines(extent = "none")

clean_table

Clinical Trial Results
Treatment	N	Response_Rate	CI_Lower	p_value
Drug A	125	72.00%	64.00%–80.00%	0.001
Drug B	130	65.00%	57.00%–73.00%	0.023
Placebo	128	45.00%	37.00%–53.00%	NA

Performance considerations

For large tables with many rows:

# Example with a larger dataset
large_data <- tibble(
  ID = 1:100,
  Group = rep(c("A", "B"), 50),
  Value1 = rnorm(100, mean = 10, sd = 2),
  Value2 = rnorm(100, mean = 20, sd = 5),
  Value3 = rnorm(100, mean = 30, sd = 8)
)

# Use opt_css for custom styling
large_data %>%
  head(20) %>%  # Show first 20 rows for example
  gt() %>%
  fmt_number(
    columns = c(Value1, Value2, Value3),
    decimals = 2
  ) %>%
  opt_css(
    css = "
    .gt_table {
      font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
    }
    .gt_col_heading {
      font-weight: 600;
    }
    .gt_row {
      padding-top: 2px;
      padding-bottom: 2px;
    }
    "
  ) %>%
  tab_options(
    container.height = px(400),
    container.overflow.y = "auto"
  )

ID	Group	Value1	Value2	Value3
1	A	7.33	22.03	42.15
2	B	12.46	22.43	38.39
3	A	9.96	16.62	20.93
4	B	11.76	12.89	23.63
5	A	9.93	21.67	34.65
6	B	10.26	21.80	25.36
7	A	13.41	24.54	32.34
8	B	11.78	25.14	30.42
9	A	11.78	19.51	29.46
10	B	10.75	17.80	45.21
11	A	8.26	25.11	35.30
12	B	8.67	13.34	35.65
13	A	8.98	20.48	24.32
14	B	9.00	21.08	22.59
15	A	7.64	12.14	16.97
16	B	9.73	17.77	44.06
17	A	11.27	19.36	43.54
18	B	9.98	23.48	29.53
19	A	7.45	22.41	41.67
20	B	10.79	31.77	30.59

Tip

For very large tables, consider using tab_options() with container.height to create a scrollable table, or paginate your results.

Conclusion

The gt package transforms R data frames into publication-quality tables with a coherent grammar-based approach. Key takeaways:

Start simple: Basic gt() creates a functional table
Layer complexity: Add formatting, styling, and annotations progressively
Think about your audience: Academic vs. business tables have different conventions
Export appropriately: Choose the right format for your medium

Resources

gt documentation: Official documentation
gtExtras: Extended functionality
gt cookbook: Practical examples

Tables deserve as much attention as plots in your data communication toolkit. With gt, you have the power to create tables that are not just functional, but beautiful.

Tip

Remember: A well-designed table can often communicate complex relationships more effectively than a plot. Choose your medium wisely.

Happy table-making!

Publication-ready tables with {gt} in R

Understanding the gt philosophy

Table anatomy

Essential formatting

Column formatting

Column labels

Advanced styling

Conditional formatting

Adding summary rows

Real-world example: Systematic review table

Integration with gtExtras

Export options

Save as image

HTML output

Word and LaTeX

Best practices

Design principles

Performance considerations

Conclusion

Resources

Let's Transform Your Data into Impact