Add color mixing plots for multi variable spatial visualization

[jiajic]

It can be helpful to view spatial patterns by overlapping multiple variables. A potential way of doing this is by independently assigning color gradients to the variables and then mixing the colors afterwards.
So far HSV #61 and RGB #62 have been added.

The HSV implementation should work for any two colors.
Additional color overlaps might be best done by mixing in the RGB space.
HSV however, does guarantee a color mixture output orthogonally to white and black, whereas RGB can produce surprisingly bright or dark colors. In both cases, color picking for the gradients is going to be important for interpretability.

Example implementation:

# load data
sl <- GiottoData::loadSubObjectMini("spatLocsObj")
cm <- GiottoData::loadSubObjectMini("cellMetaObj")
comb <- merge(sl[], cm[], by = "cell_ID")


# Create dummy gradient legends
legend_plot1 <- ggplot(comb, aes(x = sdimx, y = sdimy, color = nr_feats)) +
  geom_point() +
  scale_color_gradient(low = "black", high = "green", name = "Value 1") +
  theme_void()
legend_plot2 <- ggplot(comb, aes(x = sdimx, y = sdimy, color = sdimy)) +
  geom_point() +
  scale_color_gradient(low = "black", high = "red", name = "Value 2") +
  theme_void()


# Extract colors for legend plots
legend_colors1 <- ggplot_build(legend_plot1)$data[[1]]$colour
legend_colors2 <- ggplot_build(legend_plot2)$data[[1]]$colour

mixed_colors <- mixRGB(legend_colors1, legend_colors2)
comb$temp_colors <- mixed_colors

# Plot the actual data with color mixture
a1 <- ggplot(comb, aes(x = sdimx, y = sdimy)) +
  geom_point(aes(color = legend_colors1), size = 3, show.legend = F) +
  scale_color_identity() +
  theme_dark()

a2 <- ggplot(comb, aes(x = sdimx, y = sdimy)) +
  geom_point(aes(color = legend_colors2), size = 3, show.legend = F) +
  scale_color_identity() +
  theme_dark()

a3 <- ggplot(comb, aes(x = sdimx, y = sdimy)) +
  geom_point(aes(color = mixed_colors), size = 3, show.legend = F) +
  scale_color_identity() +
  theme_dark()

plot(a1) # var 1
plot(a2) # var 2
plot(a3) # mixture

var1

var2

mixture