Uncategorised

qu_plot_types

Every kind of plot Qu draws, one panel each.

catalog/qu_plot_types.qu

The third sheet in the reference set, after `qu_markers.qu` (every glyph) and `qu_line_styles.qu` (every line style). This one answers a different question: not "what can a series look like" but "what SHAPE of picture can I ask for", which is the thing you want to see all at once when choosing how to show a result.

Each panel is the smallest honest example of its type -- real data, real axes, no decoration beyond what the type itself needs.

Run: qu run catalog/qu_plot_types.qu

Results

printed

wrote qu_plot_types.svg and .pdf

Drawn by running the program above.

The whole program

Runnable as it stands: qu run catalog/qu_plot_types.qu

# Every kind of plot Qu draws, one panel each.

#

# The third sheet in the reference set, after `qu_markers.qu` (every glyph)

# and `qu_line_styles.qu` (every line style). This one answers a different

# question: not "what can a series look like" but "what SHAPE of picture

# can I ask for", which is the thing you want to see all at once when

# choosing how to show a result.

#

# Each panel is the smallest honest example of its type -- real data, real

# axes, no decoration beyond what the type itself needs.

#

# Run:  qu run catalog/qu_plot_types.qu


theme("publication")
figure_size(7.16, 9.0, unit = "in")

x  = linspace(0, 10, 200)
xs = linspace(0, 10, 12)

split plot 5 3

# ---- lines -------------------------------------------------------------

next plot 1
plot(x, sin(x))
title("plot")
grid on

next plot 2
scatter(xs, sin(xs))
title("scatter")
grid on

next plot 3
stem(xs, sin(xs))
title("stem")
grid on

next plot 4
stair(xs, sin(xs))
title("stair / step")
grid on

next plot 5
area(x, sin(x) + 1.5)
title("area")
grid on

next plot 6
fill_between(x, sin(x) - 0.3, sin(x) + 0.3)
plot(x, sin(x))
title("fill_between")
grid on

# ---- categorical --------------------------------------------------------

next plot 7
bar((4, 7, 3, 6))
xticklabels("A", "B", "C", "D")
title("bar")
grid on

# Stacking and grouping are done by CALLING AGAIN, not by passing several

# vectors to one call: each call adds a series to the panel, and the panel

# is what stacks or groups them. `stackbar(a, b)` reads `a` as x and `b` as

# y, exactly like `plot` -- so the one-call form silently draws a single

# series and looks like a plain `bar`.

next plot 8
bin = (1, 2, 3, 4)
stackbar(bin, (4, 7, 3, 6))
stackbar(bin, (2, 1, 4, 2))
xticklabels("A", "B", "C", "D")
title("stackbar")
grid on

next plot 9
groupbar(bin, (4, 7, 3, 6))
groupbar(bin, (2, 1, 4, 2))
xticklabels("A", "B", "C", "D")
title("groupbar")
grid on

# ---- distributions ------------------------------------------------------

next plot 10
hist(randn(2000), bins = 30)
title("hist")
grid on

next plot 11
boxplot(randn(200), randn(200) + 1.5, randn(200) - 1)
title("boxplot")
grid on

next plot 12
errorbar(xs, sin(xs), 0.15 + 0*xs)
title("errorbar")
grid on

# ---- scales and fields ---------------------------------------------------

next plot 13
f = logspace(0, 4, 200)
semilogx(f, 20*log10(1 ./ sqrt(1 + (f/100).^2)))
title("semilogx")
grid on

next plot 14
loglog(f, 1 ./ f)
title("loglog")
grid on

next plot 15
bubble(xs, sin(xs), 4 + 8*abs(sin(xs)))
title("bubble")
grid on

savefig("qu_plot_types.svg")
savefig("qu_plot_types.pdf")
print("wrote qu_plot_types.svg and .pdf")