another attempt to give examples access to data

rdboyes · rdboyes · commit ba0462205fb1 · 2026-01-11T13:50:07.000-05:00
diff --git a/data/penguins.parq b/data/penguins.parq
diff --git a/docs/examples/geoms/geom_bar.jl b/docs/examples/geoms/geom_bar.jl
@@ -38,9 +38,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # For basic counting, provide only an x aesthetic. The height of each bar will represent the count of observations in each category.
 
diff --git a/docs/examples/geoms/geom_boxplot.jl b/docs/examples/geoms/geom_boxplot.jl
@@ -48,9 +48,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # Create a basic boxplot showing the distribution of bill length for each species:
 
diff --git a/docs/examples/geoms/geom_col.jl b/docs/examples/geoms/geom_col.jl
@@ -44,10 +44,9 @@
 
 using TidierPlots
 using TidierData
-using DataFrames
-using Parquet2
+using DataFrame
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # First, we create a summarized dataset with mean bill length by species and sex:
 
diff --git a/docs/examples/geoms/geom_density.jl b/docs/examples/geoms/geom_density.jl
@@ -41,9 +41,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # Create a basic density plot showing the distribution of bill length:
 
diff --git a/docs/examples/geoms/geom_errorbars.jl b/docs/examples/geoms/geom_errorbars.jl
@@ -41,9 +41,8 @@ using TidierPlots
 using TidierData
 using DataFrames
 using Statistics
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # First, create a summary dataset with means and standard deviations:
 
diff --git a/docs/examples/geoms/geom_hist.jl b/docs/examples/geoms/geom_hist.jl
@@ -47,9 +47,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # Create a basic histogram of bill length:
 
diff --git a/docs/examples/geoms/geom_hvlines.jl b/docs/examples/geoms/geom_hvlines.jl
@@ -40,9 +40,8 @@
 using TidierPlots
 using TidierData
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # ## Horizontal Lines
 
diff --git a/docs/examples/geoms/geom_lines.jl b/docs/examples/geoms/geom_lines.jl
@@ -88,9 +88,7 @@ ggplot(df, @aes(x = x, y = y)) +
 
 # `geom_path()` connects points in the order they appear in the data. This is useful for trajectories or paths through 2D space:
 
-using Parquet2
-
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 ggplot(penguins, @aes(x = bill_length_mm, y = bill_depth_mm)) +
     geom_path(alpha = 0.3)
diff --git a/docs/examples/geoms/geom_point.jl b/docs/examples/geoms/geom_point.jl
@@ -44,9 +44,8 @@ using DataFrames
 using RDatasets
 using CategoricalArrays
 using Makie
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 diamonds = dataset("ggplot2", "diamonds");
 mtcars = dataset("datasets", "mtcars");
 
diff --git a/docs/examples/geoms/geom_rainclouds.jl b/docs/examples/geoms/geom_rainclouds.jl
@@ -43,9 +43,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 ggplot(penguins) +
 geom_rainclouds(@aes(x = species, y = bill_depth_mm))
diff --git a/docs/examples/geoms/geom_smooth.jl b/docs/examples/geoms/geom_smooth.jl
@@ -36,9 +36,9 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
+
 # ## Basic LOESS Smoothing
 
 # The default method is LOESS (locally estimated scatterplot smoothing), which creates a flexible smooth line that follows the data:
diff --git a/docs/examples/geoms/geom_text.jl b/docs/examples/geoms/geom_text.jl
@@ -47,9 +47,8 @@
 using TidierPlots
 using TidierData
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # ## Basic Text Labels
 
diff --git a/docs/examples/geoms/geom_tile.jl b/docs/examples/geoms/geom_tile.jl
@@ -86,9 +86,8 @@ ggplot(df_mandelbrot, @aes(x = x, y = y, z = z)) +
 
 
 using Statistics
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 
 # Select numeric columns and compute correlation
 numeric_cols = [:bill_length_mm, :bill_depth_mm, :flipper_length_mm, :body_mass_g]
diff --git a/docs/examples/geoms/geom_violin.jl b/docs/examples/geoms/geom_violin.jl
@@ -44,9 +44,8 @@
 
 using TidierPlots
 using DataFrames
-using Parquet2
 
-penguins = DataFrame(Parquet2.readfile(joinpath("penguins.parq")))
+penguins = penguins_data()
 # ## Basic Violin Plot
 
 # Create a basic violin plot showing the distribution of bill length for each species:
diff --git a/src/TidierPlots.jl b/src/TidierPlots.jl
@@ -161,6 +161,11 @@ export scale_size
 export scale_shape
 export scale_linewidth
 
+# example data
+
+using Parquet2
+export penguins_data
+
 # default options
 
 const plot_log = Ref{Bool}(true)
diff --git a/src/geoms/geom_bar.jl b/src/geoms/geom_bar.jl
@@ -227,6 +227,7 @@ grouping variable to `dodge` within the aesthetic.
 # Examples
 
 ```julia
+penguins = penguins_data()
 # vertical bar plot
 ggplot(penguins) + geom_bar(@aes(x = species))
 
@@ -237,7 +238,7 @@ ggplot(penguins) + geom_bar(@aes(y = species))
 ggplot(penguins, @aes(x = species, fill=sex)) + geom_bar()
 
 # dodged
-ggplot(penguins, @aes(x = species, fill=sex, dodge = sex)) + geom_bar()
+ggplot(penguins, @aes(x = species, dodge = sex)) + geom_bar()
 ```
 """
 geom_bar = geom_template("geom_bar", String[], :BarPlot;
diff --git a/src/geoms/geom_boxplot.jl b/src/geoms/geom_boxplot.jl
@@ -67,6 +67,7 @@ Compactly displays the distribution of continuous data.
 # Examples
 
 ```julia
+penguins = penguins_data()
 ggplot(penguins, @aes(x=species, y=bill_length_mm)) +
     geom_boxplot()
 
diff --git a/src/geoms/geom_density.jl b/src/geoms/geom_density.jl
@@ -31,6 +31,8 @@ Represent data as a smooth density curve.
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x=bill_length_mm)) + geom_density()
 
 ggplot(penguins, @aes(x=bill_length_mm)) +
diff --git a/src/geoms/geom_density_2d.jl b/src/geoms/geom_density_2d.jl
@@ -106,6 +106,8 @@ Represent data as a 2D density contour using 2D kernel density estimation.
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x=bill_length_mm, y=bill_depth_mm)) +
     geom_density_2d()
 
@@ -152,6 +154,8 @@ Represent data as a filled 2D density contour using 2D kernel density estimation
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x=bill_length_mm, y=bill_depth_mm)) +
     geom_density_2d_filled()
 
diff --git a/src/geoms/geom_hist.jl b/src/geoms/geom_hist.jl
@@ -32,6 +32,8 @@ Represents data as a histogram.
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x = bill_length_mm)) +
     geom_histogram()
 
diff --git a/src/geoms/geom_point.jl b/src/geoms/geom_point.jl
@@ -106,6 +106,8 @@ The biggest potential problem with a scatterplot is overplotting: whenever you h
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 p = ggplot(penguins, @aes(bill_length_mm, bill_depth_mm))
 p + geom_point()
 
diff --git a/src/geoms/geom_rainclouds.jl b/src/geoms/geom_rainclouds.jl
@@ -46,6 +46,8 @@
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins) +
     geom_rainclouds(@aes(x = species, y = bill_depth_mm/10, color = species), size = 5) +
     scale_y_continuous(labels = "{:.1f} cm") +
diff --git a/src/geoms/geom_smooth.jl b/src/geoms/geom_smooth.jl
@@ -36,6 +36,8 @@ df = DataFrame(x = xs, y = ys)
 
 ggplot(df, @aes(x = x, y = y)) + geom_smooth() + geom_point()
 
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x = bill_length_mm, y = bill_depth_mm)) +
     geom_smooth(color=:red, linewidth=10, alpha=0.5)
 ```
diff --git a/src/geoms/geom_violin.jl b/src/geoms/geom_violin.jl
@@ -33,6 +33,8 @@ Represents data as a violin plot.
 # Examples
 
 ```julia
+penguins = penguins_data()
+
 ggplot(penguins, @aes(x=species, y=bill_length_mm)) +
     geom_violin()
 
diff --git a/src/util.jl b/src/util.jl
@@ -24,3 +24,8 @@ function accepted_options_by_type(type::Symbol)
     lowercase_type = Symbol(lowercase(string(type)))
     return Makie.attribute_names(Plot{getfield(Makie, lowercase_type)})
 end
+
+function penguins_data()
+    project_path(parts...) = normpath(joinpath(@__DIR__, "..", parts...))
+    mydata = DataFrame(Parquet2.readfile(project_path("data/penguins.parq")))
+end
diff --git a/test/penguin_points.png b/test/penguin_points.png
diff --git a/test/penguins.parq b/test/penguins.parq
diff --git a/test/runtests.jl b/test/runtests.jl
@@ -26,7 +26,7 @@ using Test, TestItems, TestItemRunner
 
     # load the penguins dataset
 
-    penguins = DataFrame(Parquet2.readfile(joinpath(@__DIR__, "penguins.parq")))
+    penguins = penguins_data()
 
     # configure TidierPlots to not show the plot and not output logs
 
