CODEWITHSUNDEEP
pythonpandaspandas-styles
Ankit Goel
Ankit Goel

Reputation: 360

Applying pandas bar style to a dataframe using values from another dataframe

I have df1 and df2. I want to show bars in cells of df1 using values from df2. I was able to apply other forms of styling using below code, but with bars you cannot use this method.

def color_cells(s):
    if s > 90:
        return 'color:{0}; font-weight:bold'.format('green')
    elif s>80:
        return 'background-color: light yellow;color:{0}; font-weight:regular'.format('dark yellow')
    else:
        return 'color:{0}; font-weight:bold'.format('red')

df1.style.apply(lambda x: df2.applymap(color_cells), axis=None)

My code for getting bars in df2 is 

df2.style.bar(color=['#d65f5f', '#5fba7d'])

How do I apply above code to df1? Index and column names are same.

Adding sample dataframe:

df1=pd.DataFrame(np.random.rand(15, 10))
df2=pd.DataFrame(np.random.rand(15, 10)*100)

Upvotes: 1

Views: 967

Answers (1)

Roy2012
Roy2012

Reputation: 12503

Here's the code that does that: 

df1=pd.DataFrame(np.random.rand(15, 10))
df2=pd.DataFrame(np.random.rand(15, 10)*100)

pct = (df2 - df2.min()) / (df2.max() - df2.min() )*100

def make_bar_style(x):
    return f"background: linear-gradient(90deg,#5fba7d {x}%, transparent {x}%); width: 10em"    
pct.applymap(make_bar_style).shape

df1.style.apply(lambda x: pct.applymap(make_bar_style), axis=None)

The result is:

enter image description here

To demonstrate the fact that bar sizes are driven by df2, consider the following: 

df2 = pd.DataFrame(np.mgrid[0:15, 0:10][0])

The result is then:

enter image description here

Upvotes: 4

Related Questions