styler | Non-invasive pretty printing of R code

Instead of calling your function for each group, call it for the whole dataframe. Inside the function, use groupby(level=0).transform('mean') to get means for each group, and then compare with col > means:

The custom google maps styling must be passed to the customMapStyle prop and you need to set a specific width and height in the normal style prop of the MapView. Setting absoluteFillObject does something different than you might have thought and according to the react-native documentation there is

Currently, there is no difference between using absoluteFill vs. absoluteFillObject.

Thus, absoluteFill is for the following use case.

A very common pattern is to create overlays with position absolute and zero positioning (position: 'absolute', left: 0, right: 0, top: 0, bottom: 0), so absoluteFill can be used for convenience and to reduce duplication of these repeated styles.

This will not set a height and a width! We need to do this manually which is documented here.

The following solves your problem.

I'm not sure why the difference is, but here's a simple fix:

Possible solution is following:

OK I finally understood the issue. The problem is when the first row does not meet the criteria, the highlight_rows() returns None which is not the right dimension. To solve this, just return something of the same shape when the condition is False:

As posted at the end of my question, I managed to arrive at my desired solution with:

I changed the background: orange to background-color: orange, if you use background-color: none or simple "" in your else statement doesn't effect the output. See:

From version 1.3.0, Pandas applymap accepts a subset parameter:

subset : label, array-like, IndexSlice, optional
A valid 2d input to DataFrame.loc[], or, in the case of a 1d input or single key, to DataFrame.loc[:, ] where the columns are prioritised, to limit data to before applying the function.

So, in order, for instance, to colorize "Col1" only in the Excel output file, you can modify your code like this:

This is just a stopgap solution, but you can manually specify the usual %Y-%m-%d display format for your date column as follows: