Skip to main content
Version: 0.1.1

reactable formatters

reactablefmtr is the R package that dresses up reactable cells: data bars, color tiles and scales, pill buttons, icon sets, merged columns, highlights and bubble grids. sdvplot.reactable ports its cell formatters to reactable-py as ordinary reactable.Column objects, with reactablefmtr's semantics and defaults (the same color ramps, buckets and layouts), and adds the SportsDataverse parts: any formatter that takes a color column also takes a team column and a league, filling each cell in that team's color with text that stays readable, and data bars and merged columns can carry the row's logo, headshot or wordmark. Icons are inline SVG and tooltips are the cell's title, so the tables need no Font Awesome and no extra JavaScript.

Each formatter takes the frame you pass to Reactable (it reads ranges, buckets and team colors from it) and the column's id. Every section below is one formatter on real data from sportsdataverse-py, across the NFL, NBA, WNBA, NHL, MLB and college sports. The reactable cookbook covers the logo, headshot and team-color columns these build on.

import polars as pl
import sportsdataverse.cfb as cfb
import sportsdataverse.mbb as mbb
import sportsdataverse.mlb as mlb
import sportsdataverse.nba as nba
import sportsdataverse.nfl as nfl
import sportsdataverse.nhl as nhl
import sportsdataverse.wnba as wnba
from reactable import Column, Reactable, embed_css

import sdvplot
from sdvplot.reactable import (
reactable_sdv_bubble_grid,
reactable_sdv_color_scales,
reactable_sdv_color_tiles,
reactable_sdv_data_bars,
reactable_sdv_highlight,
reactable_sdv_icon_assign,
reactable_sdv_icon_sets,
reactable_sdv_icon_trend,
reactable_sdv_logos,
reactable_sdv_merge_column,
reactable_sdv_pill_buttons,
)

NFL_SEASON = 2025 # nflverse and cfbfastR name a season by the year it starts
SEASON = 2026 # the 2026 MLB and WNBA seasons, and the 2025-26 NBA, NHL and college basketball seasons
embed_css()

1. Data bars in team colors: NBA wins​

reactable_sdv_data_bars draws each value as a bar, its length the value's share of the column's largest. With team_col and league every bar takes its team's primary color, the number inside switches to white on a dark bar, and mark="logo" puts the team's logo at the end of the bar. tooltip=True gives each cell a hover title. The data is the 2025-26 NBA regular season, summed from ESPN team box scores of the regular-season schedule's games (the NBA Cup final, played in December, is not one).

# regular-season games only: season_type 2 also holds the NBA Cup final, which does not count in the standings
nba_schedule = nba.load_nba_schedule(seasons=[SEASON]).filter(
(pl.col("season_type") == 2) & (pl.col("type_abbreviation") == "STD")
)
nba_season = (
nba.load_nba_team_boxscore(seasons=[SEASON])
.filter(pl.col("game_id").is_in(nba_schedule["game_id"].implode()))
.group_by("team_abbreviation", "team_display_name", maintain_order=True)
.agg(gp=pl.len(), wins=pl.col("team_winner").sum())
)
assert nba_season["gp"].max() <= 82, "more games than an 82-game regular season"
nba_teams = (
nba_season.filter(pl.col("gp") > 10)
.sort(["wins", "team_abbreviation"], descending=[True, False])
.head(12)
.drop("gp")
)
Reactable(
nba_teams,
columns=[
Column(id="team_abbreviation", show=False),
Column(id="team_display_name", name="Team", min_width=190),
reactable_sdv_data_bars(
nba_teams,
team_col="team_abbreviation",
league="nba",
mark="logo",
mark_height=22,
bar_height=24,
round_edges=True,
tooltip=True,
id="wins",
name="Wins",
min_width=320,
),
],
default_page_size=12,
)

2. Color tiles: NFL offense and defense by EPA​

reactable_sdv_color_tiles fills each cell with a rounded tile on a color ramp: reactablefmtr's default blue-white-orange, or any hex colors you give, here red to green. number_fmt takes a format spec (or a function); the defensive column reverses the ramp because less allowed is better. The team column is tiled too, in each team's own color. Expected points added per play come from nflverse's weekly team stats.

weeks = nfl.load_nfl_team_stats([NFL_SEASON]).filter(pl.col("season_type") == "REG")
plays = pl.col("attempts") + pl.col("sacks_suffered") + pl.col("carries")


def epa_per_play(by):
return weeks.group_by(team=by, maintain_order=True).agg(
epa=(pl.col("passing_epa") + pl.col("rushing_epa")).sum() / plays.sum()
)


nfl_epa = (
epa_per_play(pl.col("team"))
.join(epa_per_play(pl.col("opponent_team")), on="team", suffix="_allowed")
.with_columns(net=pl.col("epa") - pl.col("epa_allowed"))
.sort("net", descending=True)
.head(16)
)
red_green = ["#d7301f", "#ffffbf", "#1a9850"]
Reactable(
nfl_epa,
columns=[
reactable_sdv_color_tiles(nfl_epa, team_col="team", league="nfl", id="team", name="Team", width=80),
reactable_sdv_color_tiles(nfl_epa, colors=red_green, number_fmt="+.3f", id="epa", name="Offense"),
reactable_sdv_color_tiles(
nfl_epa, colors=red_green[::-1], number_fmt="+.3f", id="epa_allowed", name="Defense allowed"
),
reactable_sdv_color_tiles(nfl_epa, colors=red_green, number_fmt="+.3f", bold_text=True, id="net", name="Net"),
],
default_page_size=16,
compact=True,
)

3. Color scales: the MLB standings​

reactable_sdv_color_scales is the same scale without the tile: it styles the whole cell, so the table keeps its grid. Run differential is colored on reactablefmtr's default ramp (blue at the bottom, orange at the top), and the team column carries each club's primary color at 85% opacity, its text still readable. The standings are ESPN's for the 2026 season.

mlb_table = (
mlb.espn_mlb_standings(season=SEASON)
.with_columns(pl.col("wins", "losses", "point_differential").cast(pl.Int64))
.sort(["win_percent", "team_abbreviation"], descending=[True, False])
.select(team="team_abbreviation", name="team_display_name", w="wins", l="losses", diff="point_differential")
.head(15)
)
Reactable(
mlb_table,
columns=[
reactable_sdv_color_scales(mlb_table, team_col="team", league="mlb", opacity=0.85, id="team", name=""),
Column(id="name", name="Team", min_width=180),
Column(id="w", name="W"),
Column(id="l", name="L"),
reactable_sdv_color_scales(mlb_table, number_fmt="+d", id="diff", name="Run diff."),
],
default_page_size=15,
compact=True,
)

4. Pill buttons: the WNBA's leading scorers​

reactable_sdv_pill_buttons sets each value in a rounded pill. Numbers ramp across colors; text takes the first color, or, with team_col, the row's team color. Here the team abbreviation sits in a team-colored pill and points per game ramp from gray to orange. Per-game averages come from ESPN player box scores of the 2026 regular season.

schedule = wnba.load_wnba_schedule(seasons=[SEASON]).filter(
(pl.col("season_type") == 2) & (pl.col("type_abbreviation") == "STD")
)
wnba_players = wnba.load_wnba_player_boxscore(seasons=[SEASON]).filter(
pl.col("game_id").is_in(schedule["game_id"].implode()) & ~pl.col("did_not_play")
)
scorers = (
wnba_players.sort("game_date")
.group_by("athlete_id", "athlete_display_name", maintain_order=True)
.agg(team=pl.col("team_abbreviation").last(), gp=pl.len(), ppg=pl.col("points").mean().round(1))
.filter(pl.col("gp") >= 20)
.sort(["ppg", "athlete_display_name"], descending=[True, False])
.head(12)
)
Reactable(
scorers,
columns=[
Column(id="athlete_id", show=False),
Column(id="athlete_display_name", name="Player", min_width=180),
reactable_sdv_pill_buttons(scorers, team_col="team", league="wnba", bold_text=True, id="team", name="Team"),
Column(id="gp", name="GP"),
reactable_sdv_pill_buttons(scorers, colors=["#d9d9d9", "#fa4d00"], number_fmt=".1f", id="ppg", name="PPG"),
],
default_page_size=12,
)

5. Icon sets: NHL goal differential​

reactable_sdv_icon_sets cuts the column into as many quantile buckets as icons (and as colors) and puts the bucket's icon beside each value. Three buckets of goal differential in the final 2025-26 NHL standings get a down arrow, a dash and an up arrow. The icons are inline SVG: circle, square, diamond, triangles, arrows, chevrons, star, check, x, flag, minus, plus, heart and person.

nhl_table = (
nhl.nhl_standings("2026-04-16")
.select(
team="team_abbrev_default",
name="team_name_default",
pts="points",
gf="goal_for",
ga="goal_against",
diff="goal_differential",
)
.sort(["pts", "team"], descending=[True, False])
)
Reactable(
nhl_table,
columns=[
reactable_sdv_logos(league="nhl", id="team", name="", height=24, width=60),
Column(id="name", name="Team", min_width=170),
Column(id="pts", name="PTS"),
Column(id="gf", name="GF"),
Column(id="ga", name="GA"),
reactable_sdv_icon_sets(
nhl_table,
icons=("arrow-down", "minus", "arrow-up"),
colors=("#c0392b", "#9e9e9e", "#1e8449"),
number_fmt="+d",
id="diff",
name="Goal diff.",
),
],
default_page_size=10,
compact=True,
)

6. Icon assign: SEC wins, one dot a win​

reactable_sdv_icon_assign repeats an icon up to the column's largest value and fills as many as the row's value: here each SEC team's 2025 wins as filled circles out of the most any team won, with the number to the right. With buckets= it instead fills the row's quantile bucket out of that many icons. Each row reads to a screen reader as an image labelled, for example, "10 out of 13". Wins are counted from cfbfastR's schedule of completed games.

games = cfb.load_cfb_schedule(NFL_SEASON).filter(pl.col("completed"))
results = pl.concat(
[
games.select(team_id="home_id", team="home_team", conference="home_conference", won="home_winner"),
games.select(team_id="away_id", team="away_team", conference="away_conference", won="away_winner"),
],
how="vertical_relaxed",
)
sec = (
results.filter(pl.col("conference") == "SEC")
.group_by("team_id", "team")
.agg(wins=pl.col("won").sum(), games=pl.len())
.sort(["wins", "team"], descending=[True, False])
.with_columns(logo=pl.col("team_id"))
.select("logo", "team", "wins", "games")
)
Reactable(
sec,
columns=[
reactable_sdv_logos(league="cfb", id="logo", name="", id_system="espn", height=24, width=60),
Column(id="team", name="Team", min_width=140),
reactable_sdv_icon_assign(
sec, icon_size=11, fill_color="#461d7c", show_values="right", id="wins", name="Wins", min_width=260
),
Column(id="games", name="Games"),
],
default_page_size=16,
compact=True,
)

7. Icon trend: NFL win changes, 2024 to 2025​

reactable_sdv_icon_trend marks each change with an up icon, a down icon or a dash, and colors the number to match (reactablefmtr's icon_trend_indicator). icons picks the shape: arrow, chevron or triangle. Regular-season wins come from nflverse's schedules of both seasons.

nfl_games = nfl.load_nfl_schedule([NFL_SEASON - 1, NFL_SEASON]).filter(
(pl.col("game_type") == "REG") & pl.col("result").is_not_null()
)
team_wins = (
pl.concat(
[
nfl_games.select("season", team="home_team", win=pl.col("result") > 0),
nfl_games.select("season", team="away_team", win=pl.col("result") < 0),
]
)
.group_by("team", "season")
.agg(wins=pl.col("win").sum())
.pivot("season", index="team", values="wins")
.rename({str(NFL_SEASON - 1): "before", str(NFL_SEASON): "after"})
.with_columns(change=pl.col("after") - pl.col("before"))
.sort(["change", "team"], descending=[True, False])
)
Reactable(
team_wins,
columns=[
reactable_sdv_logos(league="nfl", id="team", name="", include_name=True, height=24, min_width=100),
Column(id="before", name=str(NFL_SEASON - 1)),
Column(id="after", name=str(NFL_SEASON)),
reactable_sdv_icon_trend(
team_wins, icons="triangle", number_fmt="+d", show_zero_label=True, id="change", name="Change"
),
],
default_page_size=12,
compact=True,
)

8. Merge column with headshots: the WNBA scorers again​

reactable_sdv_merge_column shows a second column's value beside or below each value, so each player's team sits under her name in small gray text. With mark="headshot" (and mark_col naming the ESPN athlete id) the player's photo leads the cell. The merged and id columns are hidden.

scorers_named = scorers.join(sdvplot.teams("wnba").select(team="abbr", team_name="name"), on="team", how="left")
Reactable(
scorers_named,
columns=[
reactable_sdv_merge_column(
scorers_named,
merged_name="team_name",
merged_position="below",
mark="headshot",
mark_col="athlete_id",
league="wnba",
mark_height=40,
id="athlete_display_name",
name="Player",
min_width=260,
),
Column(id="athlete_id", show=False),
Column(id="team", show=False),
Column(id="team_name", show=False),
Column(id="gp", name="GP"),
Column(id="ppg", name="PPG"),
],
default_page_size=12,
)

9. Highlights: men's college basketball's best and worst units​

reactable_sdv_highlight bolds a column's maximum (green), minimum (red) or both, optionally on a highlighter. Among the top 20 men's college basketball teams of 2025-26 by adjusted efficiency margin it marks the best and worst adjusted offense; for defense, where lower is better, a (min, max) pair swaps the colors.

names = mbb.load_mbb_team_boxscore(seasons=[SEASON]).select("team_id", "team_location").unique("team_id")
mbb_ratings = mbb.load_mbb_ratings(SEASON).with_columns(pl.col("team_id").cast(pl.Int32)) # ids ship as text here
assert names.schema["team_id"] == mbb_ratings.schema["team_id"]
top20 = (
mbb_ratings.sort("rank")
.head(20)
.join(names, on="team_id", how="left")
.with_columns(pl.col("adj_em", "adj_o", "adj_d").round(1), logo=pl.col("team_id"))
.select("rank", "logo", "team_location", "adj_em", "adj_o", "adj_d")
)
Reactable(
top20,
columns=[
Column(id="rank", name="#", width=50),
reactable_sdv_logos(league="mbb", id="logo", name="", id_system="espn", height=24, width=60),
Column(id="team_location", name="Team", min_width=150),
Column(id="adj_em", name="Adj. margin"),
reactable_sdv_highlight(
top20, which="min_max", highlighter=("#fde0dd", "#e5f5e0"), id="adj_o", name="Adj. offense"
),
reactable_sdv_highlight(
top20,
which="min_max",
font_color=("green", "red"),
highlighter=("#e5f5e0", "#fde0dd"),
id="adj_d",
name="Adj. defense",
),
],
default_page_size=20,
compact=True,
)

10. Bubble grid: NHL goals for and against, in team colors​

reactable_sdv_bubble_grid draws each value as a circle (or a square, shape="squares") whose width is its share of the column's largest, up to 100 pixels, colored on a ramp or, with team_col, in the team's color. The Atlantic Division's goals for (primary colors) and against (secondary colors, as squares) make a grid to scan for the big and small numbers.

atlantic = (
nhl.nhl_standings("2026-04-16")
.filter(pl.col("division_name") == "Atlantic")
.sort("division_sequence")
.select(team="team_abbrev_default", gf="goal_for", ga="goal_against")
)
Reactable(
atlantic,
columns=[
reactable_sdv_logos(league="nhl", id="team", name="", include_name=True, height=24, min_width=100),
reactable_sdv_bubble_grid(atlantic, team_col="team", league="nhl", id="gf", name="Goals for", min_width=120),
reactable_sdv_bubble_grid(
atlantic,
team_col="team",
league="nhl",
which="secondary",
shape="squares",
id="ga",
name="Goals against",
min_width=120,
),
],
)

Run it yourself​

Download the notebook (outputs cleared) or open it on GitHub.