CODEWITHSUNDEEP
pythonpandasdataframe
Shivangi Singh
Shivangi Singh

Reputation: 11

Computing absolute difference between two dataframes

I am new to pandas and I was wondering if there was a way I could run formulas on two dataframes.

I have two dfs that have the same columns and same index. example each df looks like this

x y z a 2 4 5 b 7 9 0

I want to take df1, df2 and find the absolute difference for x ,y, z something like (df1[n]-df2[n] / df1[n]) * 100 for each n element of columns, and store the result in a new df.

Note: I just want the difference to be positive. (Misunderstood the meaning of absolute)

Upvotes: 1

Views: 5585

Answers (2)

Juan Morais
Juan Morais

Reputation: 79

If you use the formula you provided ((df1[n]-df2[n] / df1[n]) * 100), you will get different results. You can use the Pandas abs method: https://docs.python.org/3/library/functions.html#abs

Some examples:

a = {'x': [2], 'y': [4], 'z': [5]}
b = {'x': [7], 'y': [9], 'z': [0]}

df1 = pd.DataFrame(a)
df2 = pd.DataFrame(b)

df_abs = abs(df1 - df2)  # Absolute difference
df_abs = (df1 - df2 / df1) * 100  # Different results

Upvotes: 1

Kenan
Kenan

Reputation: 14124

Yes run them as you wrote them

absolute

(df1 - df2).abs()

or

(df1 -df2).div(df2) * 100

Upvotes: 1

Related Questions