将R数据清理管道（filter、select、arrange）转换为Python实现

Got it, let's translate your R dplyr pipeline step-by-step into Pandas so you can replicate the exact same logic in Python.

Complete Equivalent Code (Pandas 1.1.0+) #

This uses the cleaner key parameter in sort_values() (available in Pandas 1.1.0 and later):

import pandas as pd

# Match your R workflow exactly
df = (dataframe
      # Filter rows where 'Sheet1' contains 'abc'
      .loc[dataframe['Sheet1'].str.contains('abc', na=False)]
      # Select only the 'product' column
      [['product']]
      # Sort by the length of the 'product' string
      .sort_values(by='product', key=lambda x: x.str.len()))

Step-by-Step Breakdown #

Let's map each part of your R code to Pandas:

• Filter rows with 'abc' in 'Sheet1':
  R's filter(grepl('abc', Sheet1)) becomes dataframe['Sheet1'].str.contains('abc', na=False) in Pandas. We wrap this in .loc[] to select matching rows. The na=False ensures any rows with missing values in 'Sheet1' are excluded, matching how grepl() handles NA values in R.
• Select the 'product' column:
  R's select(product) is as simple as [['product']] in Pandas (this preserves the DataFrame structure, rather than returning a Series). Alternatively, you can use .filter(items=['product']) if you prefer a method-chaining style closer to dplyr.
• Sort by string length of 'product':
  R's arrange(nchar(product)) translates to sort_values(by='product', key=lambda x: x.str.len()). The key parameter lets us pass a function that generates the sort key—here, we calculate the length of each string in the 'product' column, just like nchar() does in R.

For Older Pandas Versions (Pre-1.1.0) #

If you're stuck on a Pandas version without the key parameter, you can create a temporary length column, sort, then drop it:

df = (dataframe
      .loc[dataframe['Sheet1'].str.contains('abc', na=False)]
      [['product']]
      .assign(product_length=lambda x: x['product'].str.len())
      .sort_values(by='product_length')
      .drop(columns='product_length'))

内容的提问来源于stack exchange，提问作者Catherine Zhang