pandas has predefined formatting rules written and stored as functions that can be readily used.

The highlight_null method highlights all NaNs or Null values in the data with a specified color. In the DataFrame under discussion, the Age and Cabin columns have NaNs. Hence, in the following screenshot, the NaNs are flagged in blue in these columns.

The following snippet highlights the NaN values in these columns:

df.style.highlight_null(null_color = "blue")

This results in the following output:

The highlight_max and highlight_min methods apply highlighting (with a chosen color) to the maximum or minimum value across either axis. In the following example, the minimum values in each column have been highlighted:

 df.iloc[0:10, :].style.highlight_max(axis = 0)

Please note that only columns with the numeric datatype are subject to highlighting.

The following screenshot highlights the maximum values for each column:

In the preceding code, highlight_max has been used to highlight the maximum values in each column.

Next, we use the same function to find the maximum for each column, changing the value of the axis parameter while doing so:

df.style.highlight_max(axis = 1)

The following screenshot shows the highlighted maximum values across columns: 

Now, let's use the highlight_min function to highlight the minimum values with a custom-defined color. Both highlight_min and highlight_max have the same syntax and accept the same set of parameters:

A background color gradient based on conditional formatting can be applied to columns to give a sense of high, medium, and low values based on color. The backgrounds are colored with different colors based on whether they are high, medium, or low.

The background gradient of the table can be controlled through the background_gradient() styling function. Any existing colormaps or user-defined colormaps can be used as a gradient. Parameters such as low and high help us use part of the colormap's color range. Further, the axis and subset parameters can be set to vary the gradient along a certain axis and subset of columns:

df.style.background_gradient(cmap='plasma', low = 0.25, high = 0.5) 

This results in the following output:

Styling can also be done independently of values. Let's modify the properties to change the font color, background color, and border color. You can do so by using the following code.

df.style.set_properties(**{'background-color': 'teal',
                               'color': 'white',
                               'border-color': 'black'})

This results in the following output:

Styling options also help us control the numerical precision. Consider the following DataFrames:

Take a look at the following code, which sets the precision to 2 decimal places or rounds off a number number to 2 decimal places.

rand_df.style.set_precision(2)

This results in the following output:

Now, let's set a caption for the preceding DataFrame:

rand_df.style.set_precision(2).set_caption("Styling Dataframe : Precision Control")

This results in the following output:

df.style.set_table_styles([{'selector': 'tr:hover','props': [('background-color', 'lawngreen')]}]

This results in the following output:

The hide_index and hide_columns styling options allow us to hide either the index or specified columns when they're displayed. In the following code, we have hidden the default index column:

df.style.hide_index()

The following screenshot shows the output DataFrame, without its index:

Now, let's use the hide_columns option to hide the "Name", "Sex", "Ticket", and "Cabin" columns:

df.style.hide_columns(["Name", "Sex", "Ticket", "Cabin"])

The following screenshot displays the columns that are shown after hiding a few columns from a DataFrame: