如何在Python中计算Pandas DataFrame得分列0到100步长为1的百分位数

Hey there! Glad you asked—this is a super common task, and Pandas makes it just as straightforward as R does. Let’s break down how to replicate that R code in Python.

Basic Implementation #

You’ll use Pandas’ built-in quantile() method for your DataFrame column, paired with NumPy’s linspace() to generate the sequence of probability values (just like R’s seq(0, 1, 0.01)).

Here’s the code:

import pandas as pd
import numpy as np

# Assuming your DataFrame is named 'df' with a 'score' column
perc = df['score'].quantile(q=np.linspace(0, 1, 101))

Let’s unpack this:

• np.linspace(0, 1, 101) generates 101 evenly spaced values from 0 to 1 (inclusive), which exactly matches seq(0, 1, 0.01) in R (0.00, 0.01, 0.02, ..., 1.00).
• The q parameter in Pandas’ quantile() corresponds directly to R’s probs parameter—it defines which percentiles to calculate.

Alternative Without NumPy #

If you don’t want to rely on NumPy, you can create the probability sequence with a simple list comprehension:

# Generate 0.00 to 1.00 in 0.01 increments
probs = [x / 100 for x in range(0, 101)]
perc = df['score'].quantile(q=probs)

Matching R’s Default Behavior (Optional) #

One quick note: Pandas and R use slightly different default methods for calculating quantiles. If you need your results to perfectly match R’s default quantile() output (which uses type 7), you can specify the method parameter in Pandas:

perc = df['score'].quantile(q=np.linspace(0, 1, 101), method='median_unbiased')

This aligns Pandas’ calculation logic with R’s default type.

That’s all! You’ll end up with a Series where each index is the probability (0.00 to 1.00) and the value is the corresponding percentile for your score column—exactly like the perc object in R.

内容的提问来源于stack exchange，提问作者red_quark