Finding error difference row wise between two dataframes in python

Question

Is there an easier way to find the percent difference between two dataframes.

For example:

df1((row1,col1) -df2(row1, col1))/average(df1(row1,col1), df2(row1,col1))

The picture shows the original dataframes, where I do it in a more manual way.

Answer 1

You can calculate the element-wise difference between two data frames like this:

diff_df = df1 - df2

The same way, you can add them together and divide them by 2. And multiply them by 100:

avg_df = (df1 + df2) / 2

You can divide diff_df by avg_df using .div() . Multiplying that with 100 should get you what you need:

diff_df / avg_df * 100

---

You can also use the pandas methods to do this:

diff_df = df1.subtract(df2)
avg_df = df1.add(df2) / 2

diff_df.div(avg_df) * 100

---

Or, as a one-liner:

df1.subtract(df2).div(df1.add(df2).div(2)).mul(100)