极客教程如何在Python中分块处理Excel文件数据？
阅读更多：Python 教程
介绍
看起来世界是由Excel统治的。在我的数据工程工作中，我惊讶地发现有多少同事在使用Excel作为决策的关键工具。虽然我不是MS Office和它们的Excel电子表格的粉丝，但我仍然会向您展示一种有效处理大型Excel电子表格的巧妙技巧。
怎么做…
在我们直接进入程序之前，让我们先了解一下使用Pandas处理Excel电子表格的一些基础知识。
1. 安装。请先安装openpyxl和xlwt。如果您不确定是否已安装，请在Python终端中使用pip freeze或pip list查看可用包。
我们将首先通过传递元组数据来创建一个Excel电子表格。然后我们将把数据加载到pandas数据框中。最后，我们将把数据框数据写入新工作簿。
import xlsxwriter
import pandas as pd
2.创建一个带有小数据的Excel电子表格。我们将拥有一个小函数，将字典数据写入Excel电子表格。所有的代码逻辑都在每个步骤中定义。
# Function : write_data_to_files
def write_data_to_files(inp_data, inp_file_name):
"""
function : create a csv file with the data passed to this code
args : inp_data : tuple data to be written to the target file
file_name : target file name to store the data
return : none
assumption : File to be created and this code are in same directory.
"""
print(f" *** Writing the data to - {inp_file_name}")
# Create a Workbook.
workbook = xlsxwriter.Workbook(inp_file_name)
# add a worksheet.
worksheet = workbook.add_worksheet()
# Start from the first cell. Rows and columns are zero indexed.
row = 0
col = 0
# read the input data and write them in rows and columns
for player, titles in inp_data:
worksheet.write(row, col, player)
worksheet.write(row, col + 1, titles)
row += 1
# close the workbook.
workbook.close()
print(f" *** Completed writing the data to - {inp_file_name}")
# Function : excel_functions_with_pandas
def excel_functions_with_pandas(inp_file_name):
"""
function : Quick overview of functions you can apply on excel with pandas
args : inp_file_name : input excel spread sheet.
return : none
assumption : Input excel spreadsheet and this code are in same directory.
"""
data = pd.read_excel(inp_file_name)
# print top 2 rows
print(f" *** Displaying top 2 rows of - {inp_file_name} \n {data.head()} ")
# look at the data types
print(f" *** Displaying info about {inp_file_name} - {data.info()}")
# Create a new spreadsheet "Sheet2" and write data into it.
new_players_info = pd.DataFrame(data=[
{"players": "new Roger Federer", "titles": 20},
{"players": "new Rafael Nadal", "titles": 20},
{"players": "new Novak Djokovic", "titles": 17},
{"players": "new Andy Murray", "titles": 3}], columns=["players", "titles"])
new_data = pd.ExcelWriter(inp_file_name)
new_players_info.to_excel(new_data, sheet_name="Sheet2")
if __name__ == '__main__':
# Define your file name and data
file_name = "temporary_file.xlsx"
# tuple data for storage
file_data = (['player', 'titles'], ['Federer', 20], ['Nadal', 20], ['Djokovic', 17], ['Murray', 3])
# write the file_data to file_name
# write_data_to_files(file_data, file_name)
# # Read excel file into pandas and apply functions.
# excel_functions_with_pandas(file_name)
if__name__ == '__main__':
# Define your file name and data
file_name = "temporary_file.xlsx"
# tuple data for storage
file_data = (['player', 'titles'], ['Federer', 20], ['Nadal', 20], ['Djokovic', 17], ['Murray', 3])
# write the file_data to file_name
# write_data_to_files(file_data, file_name)
# # Read excel file into pandas and apply functions.
# excel_functions_with_pandas(file_name)
输出
*** 将数据写入临时文件.xlsx
*** 已成功将数据写入临时文件.xlsx
*** 显示临时文件.xlsx的前2行
player titles
0 Federer 20
1 Nadal 20
2 Djokovic 17
3 Murray 3
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4 entries, 0 to 3
Data columns (total 2 columns):
# 列名 非空数量 数据类型
--- ------ -------------- -----
0 player 4 non-null object
1 titles 4 non-null int64
dtypes: int64(1), object(1)
memory usage: 192.0+ bytes
*** 显示临时文件.xlsx的信息 - None
现在，在处理大型 CSV 文件时，我们有很多选项，包括分块处理它们，但是对于 Excel 电子表格，Pandas 默认不提供分块选项。
因此，下面的程序如果您希望以块处理 Excel 电子表格将非常有帮助。
示例
def global_excel_to_db_chunks(file_name, nrows):
"""
function : handle excel spreadsheets in chunks
args : inp_file_name : input excel spread sheet.
return : none
assumption : Input excel spreadsheet and this code are in same directory.
"""
chunks = []
i_chunk = 0
# The first row is the header. We have already read it, so we skip it.
skiprows = 1
df_header = pd.read_excel(file_name, nrows=1)
while True:
df_chunk = pd.read_excel(
file_name, nrows=nrows, skiprows=skiprows, header=None)
skiprows += nrows
# When there is no data, we know we can break out of the loop.
if not df_chunk.shape[0]:
break
else:
print(
f" ** Reading chunk number {i_chunk} with {df_chunk.shape[0]} Rows")
# print(f" *** Reading chunk {i_chunk} ({df_chunk.shape[0]} rows)")
chunks.append(df_chunk)
i_chunk += 1
df_chunks = pd.concat(chunks)
# Rename the columns to concatenate the chunks with the header.
columns = {i: col for i, col in enumerate(df_header.columns.tolist())}
df_chunks.rename(columns=columns, inplace=True)
df = pd.concat([df_header, df_chunks])
print(f' *** Reading is Completed in chunks...')
if __name__ == '__main__':
print(f" *** Gathering & Displaying Stats on the excel spreadsheet ***")
file_name = 'Sample-sales-data-excel.xls'
stats = pd.read_excel(file_name)
print(f" ** Total rows in the spreadsheet are - {len(stats.index)} Rows")
# process the excel file in chunks of 1000 rows at a time.
global_excel_to_db_chunks(file_name, 1000)
*** Gathering & Displaying Stats on the excel spreadsheet ***
** Total rows in the spreadsheet are - 9994 Rows
** Reading chunk number 0 with 1000 Rows
** Reading chunk number 1 with 1000 Rows
** Reading chunk number 2 with 1000 Rows
** Reading chunk number 3 with 1000 Rows
** Reading chunk number 4 with 1000 Rows
** Reading chunk number 5 with 1000 Rows
** Reading chunk number 6 with 1000 Rows
** Reading chunk number 7 with 1000 Rows
** Reading chunk number 8 with 1000 Rows
** Reading chunk number 9 with 994 Rows
*** Reading is Completed in chunks...
输出
*** Gathering & Displaying Stats on the excel spreadsheet ***
** Total rows in the spreadsheet are - 9994 Rows
** Reading chunk number 0 with 1000 Rows
** Reading chunk number 1 with 1000 Rows
** Reading chunk number 2 with 1000 Rows
** Reading chunk number 3 with 1000 Rows
** Reading chunk number 4 with 1000 Rows
** Reading chunk number 5 with 1000 Rows
** Reading chunk number 6 with 1000 Rows
** Reading chunk number 7 with 1000 Rows
** Reading chunk number 8 with 1000 Rows
** Reading chunk number 9 with 994 Rows
*** Reading is Completed in chunks...