Read merged cells in Excel with Python

Question

I am trying to read merged cells of Excel with Python using xlrd.

My Excel: (note that the first column is merged across the three rows)

    A   B   C
  +---+---+----+
1 | 2 | 0 | 30 |
  +   +---+----+
2 |   | 1 | 20 |
  +   +---+----+
3 |   | 5 | 52 |
  +---+---+----+

I would like to read the third line of the first column as equal to 2 in this example, but it returns ''. Do you have any idea how to get to the value of the merged cell?

My code:

all_data = [[]]
excel = xlrd.open_workbook(excel_dir+ excel_file)
sheet_0 = excel.sheet_by_index(0) # Open the first tab

for row_index in range(sheet_0.nrows):
    row= ""
    for col_index in range(sheet_0.ncols):
        value = sheet_0.cell(rowx=row_index,colx=col_index).value             
        row += "{0} ".format(value)
        split_row = row.split()   
    all_data.append(split_row)

What I get:

'2', '0', '30'
'1', '20'
'5', '52'

What I would like to get:

'2', '0', '30'
'2', '1', '20'
'2', '5', '52'

Solution 1

I just tried this and it seems to work for your sample data:

all_data = []
excel = xlrd.open_workbook(excel_dir+ excel_file)
sheet_0 = excel.sheet_by_index(0) # Open the first tab

prev_row = [None for i in range(sheet_0.ncols)]
for row_index in range(sheet_0.nrows):
    row= []
    for col_index in range(sheet_0.ncols):
        value = sheet_0.cell(rowx=row_index,colx=col_index).value
        if len(value) == 0:
            value = prev_row[col_index]
        row.append(value)
    prev_row = row
    all_data.append(row)

returning

[['2', '0', '30'], ['2', '1', '20'], ['2', '5', '52']]

It keeps track of the values from the previous row and uses them if the corresponding value from the current row is empty.

Note that the above code does not check if a given cell is actually part of a merged set of cells, so it could possibly duplicate previous values in cases where the cell should really be empty. Still, it might be of some help.

Additional information:

I subsequently found a documentation page that talks about a merged_cells attribute that one can use to determine the cells that are included in various ranges of merged cells. The documentation says that it is "New in version 0.6.1", but when i tried to use it with xlrd-0.9.3 as installed by pip I got the error

NotImplementedError: formatting_info=True not yet implemented

I'm not particularly inclined to start chasing down different versions of xlrd to test the merged_cells feature, but perhaps you might be interested in doing so if the above code is insufficient for your needs and you encounter the same error that I did with formatting_info=True.

Solution 2

You can also try using fillna method available in pandas https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.fillna.html

df = pd.read_excel(dir+filename,header=1)
df[ColName] = df[ColName].fillna(method='ffill')

This should replace the cell's value with the previous value

Solution 3

For those who are looking for handling merged cell, the way OP has asked, while not overwriting non merged empty cells.

Based on OP's code and additional information given by @gordthompson's answers and @stavinsky's comment, The following code will work for excel files (xls, xlsx), it will read excel file's first sheet as a dataframe. For each merged cell, it will replicate that merged cell content over all the cells this merged cell represent, as asked by original poster.Note that merged_cell feature of xlrd for 'xls' file will only work if 'formatting_info' parameter is passed while opening workbook.

import pandas as pd
filepath = excel_dir+ excel_file
if excel_file.endswith('xlsx'):
    excel = pd.ExcelFile(xlrd.open_workbook(filepath), engine='xlrd')
elif excel_file.endswith('xls'):
    excel = pd.ExcelFile(xlrd.open_workbook(filepath, formatting_info=True), engine='xlrd')
else:
    print("don't yet know how to handle other excel file formats")
sheet_0 = excel.sheet_by_index(0) # Open the first tab
df = xls.parse(0, header=None) #read the first tab as a datframe

for e in sheet_0.merged_cells:
    rl,rh,cl,ch = e
    print e
    base_value = sheet1.cell_value(rl, cl)
    print base_value
    df.iloc[rl:rh,cl:ch] = base_value

Solution 4

Using XLRDs merged cells

ExcelFile = pd.read_excel("Excel_File.xlsx")
xl = xlrd.open_workbook("Excel_File.xlsx")
FirstSheet = xl.sheet_by_index(0)
for crange in FirstSheet.merged_cells:
    rlo, rhi,clo, chi = crange
    for rowx in range(rlo,rhi):
        for colx in range(clo,chi):
            value = FirstSheet.cell(rowx,colx).value
        if len(value) == 0:
            ExcelFile.iloc[rowx-1,colx] = FirstSheet.cell(rlo,clo).value

View more solutions

Comments

[John Y]

Even more info in this mailing list thread. formatting_info is unsupported for .xlsx files, unfortunately.

[Stavinsky]

This will fill all na cells. Even if it is not merged

[Stavinsky]

formatting_info doesn't needed for xlsx as I can see.

[jimmy]

worked for me with a minor change:- had to decrement rl and rh by 1 in the last line