Using pandas applymap() with multiple mapping functions

Question

I am using df.to_excel() to output data from a pandas dataframe to excel. To improve readability, I am using df.style.applymap() to change the color of the cell based on the contents.

Imagine I have a dataframe that looks like:

df = 
       Account  Revenue   AccountAge
0     "Boeing"     5000          5.6
1   "Lockheed"   -10000          1.2
2     "Airbus"    12000          0.6
3   "Northrop"    -3000          8.4

Where account age represents how long I have had this account on my books. Perhaps I want to flag negative revenue cells as red. I can do:

def color_neg_revenue(val):
    if val < 0:
        color = 'red'
    return "background-color: %s" % color

df = df.style.applymap(color_neg_revenue, subset=["Revenue"])

I can export this to excel, and the format looks great! But suppose I also want to flag new accounts in yellow, when I try:

def color_new_account(val):
    if val < 3:
        color = 'yellow'
    return "background-color: %s" % color

df = df.style.applymap(color_new_account, subset=["AccountAge"])

I get:

AttributeError: 'Styler' object has no attribute 'style'

Why can I not use applymap() on a dataframe that I have already used applymap() on? How can I get around this?

Answer 1

The problem is not with the function. In your first call you perform:

df = df.style.applymap(color_neg_revenue, subset=["Revenue"])

So now df is no longer a DataFrame, it is a Styler. Such Styler has of course no .style attribute anymore. You thus better assign it to another variable, and chain on that one. Like:

s = df.style.applymap(color_neg_revenue, subset=["Revenue"])

# Look mom, no I have no .style
s = s.applymap(color_new_account, subset=["AccountAge"])

You thus can chain such applymap calls when you use a Styler, but such styler has no .style attribute itself.