What is the performance impact of non-unique indexes in pandas?

When index is unique, pandas use a hashtable to map key to value O(1). When index is non-unique and sorted, pandas use binary search O(logN), when index is random ordered pandas need to check all the keys in the index O(N).

You can call sort_index method:

import numpy as np
import pandas as pd
x = np.random.randint(0, 200, 10**6)
df1 = pd.DataFrame({'x':x})
df2 = df1.set_index('x', drop=False)
df3 = df2.sort_index()
%timeit df1.loc[100]
%timeit df2.loc[100]
%timeit df3.loc[100]

result:

10000 loops, best of 3: 71.2 µs per loop
10 loops, best of 3: 38.9 ms per loop
10000 loops, best of 3: 134 µs per loop

@HYRY said it well, but nothing says it quite like a colourful graph with timings.

Plots were generated using perfplot. Code, for your reference:

import pandas as pd
import perfplot

_rnd = np.random.RandomState(42)

def make_data(n):    
    x = _rnd.randint(0, 200, n)
    df1 = pd.DataFrame({'x':x})
    df2 = df1.set_index('x', drop=False)
    df3 = df2.sort_index()

    return df1, df2, df3

perfplot.show(
    setup=lambda n: make_data(n),
    kernels=[
        lambda dfs: dfs[0].loc[100],
        lambda dfs: dfs[1].loc[100],        
        lambda dfs: dfs[2].loc[100],
    ],
    labels=['Unique index', 'Non-unique, unsorted index', 'Non-unique, sorted index'],
    n_range=[2 ** k for k in range(8, 23)],
    xlabel='N',
    logx=True,
    logy=True,
    equality_check=False)