Python will print list value but says list is empty when trying to access values (using .pop() or indexing[0] etc.)

Question

Background

I am reading in data from a .xlsx spreadsheet using Pandas 1.3.4 and Python 3.9.6 .

I want data for each item based on a code in the 'Code' column so I am first isolating a row in the dataframe ( df ) using:

row = df.loc[df['Code'] == code]

To then get individual values in the row I am accessing them using indexing and their headers as follows:

barcode = row['Barcode'].values.tolist()

I am using .values as I otherwise get a cumbersome line number and datatype output. I am using .tolist() so that I can effectively access the actual barcode rather than a scientific notation interpretation. Note that this means barcode should be a list with a single value.

Problem

Here is my problem, when I print barcode I get a sensible output, eg: [72934728491] (note this is not an actual barcode, but an example 11 digits)

Update: len(barcode) returns 1 and type(barcode) returns <class 'numpy.ndarray'> or <class 'list'> if I've used .tolist()

However, if I try to access the value inside the list using barcode[0] or barcode.pop() I get errors saying that the index is out of range or that .pop() cannot be used on an empty list.

Bizarrely, if I write a for loop to get the value out of the list it works just fine:

for item in barcode:
    print(item)

Question

Please can someone explain why this is happening and how to actually fix it (rather than having to write a for loop for every single column value I need to access on the row).

Update (More code as requested)

I have tried to construct a Minumum Viable Example to help, but am struggling to replicate the problem in other code for now, this is the section where I dissect the dataframe, unfortunately I cannot show the section where I read in the df as it contains private info, but the spreadsheet reads into the df 'normally':

for code in new_items:
    row = df.loc[df['Code'] == code] # row for a single item code
    product_display_name = row['Product Display Name'].values
    product_type = row['Product Type'].values
    description = row['Description'].values
    artist = row['Artist'].values
    barcode = row['Barcode'].values.tolist()
    finish = row['Finish'].values
    unit = row['Unit'].values
    country_of_origin = row['Country of Origin'].values
    job_number = row['Job No.'].values
    samples_date = row['Samples Date'].values

print(barcode.pop()) # this does not work
for item in barcode:
    print(item) # this does work

Answer 1

The problem turned out to be to do with pandas ' handling of empty cell NaN values.

While dataframes display NaN , somewhere in the .values and .tolist() conversion I was losing that info and having empty arrays. This was further complicated by some strange spreadsheet formatting that meant the dataframe had a 0th row of all NaN values so my loop never even hit a valid value before failing.