dbig taxa is putting the family at the end of every genus name, not just the ambiguous ones

Code is in Florian's project

Include an option in plot_tax_bar() to put the legend at the bottom instead of on the side.

Fix how the cutoff is used in make_phy_df() so that it conducts the cutoff after glomming the taxa.

• stop using aes_string_(), which is deprecated. Update to whatever replaced it.
• include a cutoff argument in the function itself so that people don't have to re-run make_tax_df() to change their cutoff
• provide an option to limit the number of taxa shown in the legend without removing them from the plot

Hi Jake,

I am getting the following error:

"Error in FUN(X[[i]], ...) : object 'extraction' not found"

when I attempt to run plot_taxon_violin

I think this is because the function was originally used to plot differences between extraction methods maybe? Is there any way to allow for specifying the x axis in the function? (similar to how plot_taxa_bar allows specifying 'sample'. In the meantime I've just changed the function on my end to whatever I would like the X axis to be..

Great stuff, really finding the package helpful.

I think this will screw things up for group-mean taxa barcharts when the cutoff is >0 because the non-zero total will be removed from the denominator. Need to make sure I understand exactly what this is doing.

# Set all counts < the cutoff to zero
phyloseq::otu_table(phyl_glommed)[phyloseq::otu_table(phyl_glommed) <
                                      cutoff] = 0

I want to remove all instances of deprecated underscore functions and replace it with the new syntax. I just feel like I should wait a bit to be more sure that users will have the newer version of rlang. Otherwise I have to implement version checks and make it backward compatible.

Hi,

When I attempt to use make_phy_df from exploratory_graphics.R, the following error is returned:

Error in group_by_(., .dots = metacols) :
could not find function "group_by_"

Seems like the function was removed?

Thanks in advance for help with this and please let me know if further information is required

Fix ordering of colour vectors so that they behave the way a user expects: The most abundant taxon should be the first colour in the user-provided vector, and so on.

Write functions to pull preddicteds and CIs out of (g)lm and (g)lmer model objects for plotting.

All the functions that deal with taxon names are slow because that object is slow. Re-implement using data frames.

Sometimes different genera in different families have the same name. Write a function to append the family name to the genus name. Code is in the Steinberg Metformin project.

make_phy_df() does weird things if the table isn't relative abundance. Check that and do something sensible.

It's past time to have a test suite for all exported functions.

group_by_() is deprecated. Figure out how to use group_by() programmatically instead. Probably merge this issue with the rlang issue.

When the leglen option is used in plot_tax_bar(), the Other category is no longer grey, but is taken from the colour vector.