How can I apply a custom mask to allow dynamic formatting

[JfosterBGC]

Hey All, not sure if this is the right place for this question, is there a way to apply a custom mask to the GT for more dynamic formatting?

I.e. want to highlight particular cells in the table based on the value of cells and data both inside and outside of the table (i.e. from parts used to assemble the table), but not based on the value of the cell itself.

According to the Documentation, it appears that I can utilise the mask functionality. Wrote some code to generate a Boolean polar table that corresponds to my table, but I can't seem to figure out how to implement it. Any suggestions for how to create some dynamic formatting? I also tried to see if I could use eval and just brute force the formattin,g but GT doesn't like eval.

[jrycw]

@JfosterBGC, I believe this is the right place to share your experience.

I'm sorry, but I’m not sure I fully understand the issue. It would be great if you could provide a small example showing the input and output, and I’ll see if I can help.

[JfosterBGC]

I have cut down my table to the more relevant parts, but what I what i have is a large dataframe (was pandas, but then converted to polar for compatibility with GT) that contains results for the same sample (Slip surface ID) in a number of different analysis (Spanners under 'Factor of Safety' with sub-analysis). I then go through that large dataframe and identify which samples meet particular conditions ('Failure modes')
For each of these different groupings I find the lowest values for that particular group and how it compares to the other analysis, i.e. I will fine the 3 lowest FoS for the 2D, EE_BRW2_ESA, for the 0_200_SE Gully 3-4 (Row group) that meet the crest Failure mode. I then stack with on top with all other analysis so I can compare all the data in one place and see how it changes.

however what I loose in the stack is an easy wat to show which analysis the data is filtered by and where we might have samples that are critical in multiple cases.

I am able to easily retain the slip surface id's that are crtitical to each analysis so my hope was to have two formats of Boxes, one showing the slipe surface critical to that analysis and ones showing were they are crictal in other analysis. something like below

I was trying to do it all in one step because I didn't know how to append functions to the GT, but I have since realized that I can use just set the Table variable equal to its self and then add more functions(See code snippet below), so i may just go that root and go column by column. But I am happy to hear suggestions on a cleaner way.

` TBL = (
GT(Sub_pl)
# Built tabel componets
.tab_header(title=md(kwargs['title']+f' for {L1}'),
subtitle=md(Subtitle))
...
...
for k, v in FoS_Analysis_Cols.items():
TBL = TBL.tab_spanner(label= md(f'{k}'), columns=v,level = 0)

`

[jrycw]

Thanks for the explanation. I think I now have a rough idea of your work.
Here are my general suggestions:
(1) If you’re performing conditional branching on a per-column basis, the mask= parameter in loc.body() might be helpful.
(2) If you need to call a function provided by Great Tables multiple times, the technique you used in the code snippet seems correct.
(3) If you need to call a custom function multiple times, the GT.pipe() method might be helpful.

[JfosterBGC]

Thanks I will look into these functions and see how they can help. I appreciate you taking the time to look at my question!

[JfosterBGC]

For those Looking back on this, I was able to solve it by doing the following:

First I went back into table building function and added a "From_Analysis" Column to the dataframe so I explicitly had a list of which analysis the slip surface was critical for.

then when in it came to the plot i did the following:

# Format the Critical FoS values in box for Col in [c for c in FoS_not_empty if not 'FoS_Ratio' in c]: ssID = Sub_df['Slip_Surface_ID'][Sub_df['From_Analysis'].apply(lambda x:Col in x)].to_list() idxs = Sub_df[Sub_df['From_Analysis'].apply(lambda x:Col in x)].index.to_list() TBL = TBL.tab_style(style = style.borders(sides="all", color = "#FF000087", style='solid', weight=3), locations=loc.body(columns= Col, rows = pl.col('Slip_Surface_ID').is_in(ssID))) TBL = TBL.tab_style(style = style.borders(sides="all", color = "#FF000087", style='solid', weight=9), locations=loc.body(columns= Col, rows = pl.col('index').is_in(idxs)))

in this block I loop through each of the analsis columns (Col) one at a time.
I then get the slip surface ID (ssID) and index (Indxs) for all the rows that have the 'Col' in question in the list. I used the pandas version of my DataFrame as I am less familar with polars and just couldn't seem to get it to work.
I then used the ssID and indxs to apply the formating, doing the Indexs second so they override the more general flagging of the ssID.

I hope others find this useful :)

[jrycw]

Slightly reformat the provided code as shown below:

# Format the Critical FoS values in box
for Col in [c for c in FoS_not_empty if not "FoS_Ratio" in c]:
    ssID = Sub_df["Slip_Surface_ID"][Sub_df["From_Analysis"].apply(lambda x: Col in x)].to_list()
    idxs = Sub_df[Sub_df["From_Analysis"].apply(lambda x: Col in x)].index.to_list()
    TBL = TBL.tab_style(
        style=style.borders(sides="all", color="#FF000087", style="solid", weight=3),
        locations=loc.body(columns=Col, rows=pl.col("Slip_Surface_ID").is_in(ssID)),
    )
    TBL = TBL.tab_style(
        style=style.borders(sides="all", color="#FF000087", style="solid", weight=9),
        locations=loc.body(columns=Col, rows=pl.col("index").is_in(idxs)),
    )