reduction operation 'argmax' not allowed for this dtype

Question

How to Take index of Last Cell in ColumnB?

I will iterate through all rows, by taking values in there. For that iteration i wanted to take index value of max row in particular column with pandas in python. df['columnNameB'].idxmax()

I also tried using axis=1 or axis='columns gave other strange errors.

idxmax() works in Series of pandas and df['colNameB'] already gives as a series of columnB values, then why idxmax() does not work properly?

code is below;

#reading from csv file url-s
def readCSV(path_csv):
    df=pd.read_csv(path_csv)
    return df

fileCSV=readCSV(r'C:\Users\Admin\Downloads\urls.csv')
print(fileCSV['columnURLnames'].idxmax())
"""for i in range(len(fileCSV.columns(1))):
    xUrl=fileCSV.iloc[0,i]
    print(xUrl)
"""

edit:

   columnA    columnB                   columnC
0   book      www.amazon.com=page?1..   onstock
1   headphone www.amazon.com=page?1..   outstock
2   NaN       www.alibaba.com..         NaN
3   NaN       www.amazon....            NaN
4   lightbulb    NaN                    inStock

I need to get number 3. Because index of max value in columnB is 3

solved: from SO

length_of_column_urls=fileCSV['columnURLnames'].last_valid_index()

for i in range(0, length_of_column_urls+1 ):

    xUrl=fileCSV.iloc[i,1]
    print(xUrl)

WHY is row 3 the the "max" value for column B? Maximum based on what criteria? — Tim Roberts, Oct 26 '22 at 21:49
Got an answer from [SO](https://stackoverflow.com/questions/15862034/access-index-of-last-element-in-data-frame) post. `fileCSV['columnB'].index[-1]` — xlmaster, Oct 27 '22 at 01:57
In fairness, that's an answer to a question quite different from the one you asked. — Tim Roberts, Oct 27 '22 at 03:24
yeah, that is not be an answer. it gives 4 not three. it counts all columns — xlmaster, Oct 27 '22 at 05:45
@TimRoberts answer should be last_valid_index() method from https://stackoverflow.com/questions/15862034/access-index-of-last-element-in-data-frame?noredirect=1&lq=1 — xlmaster, Oct 27 '22 at 06:40

bitflip · Answer 1 · 2022-10-26T21:00:17.967

1

idxmax returns the index of the highest value. There is no such value for a series of url names (strings).

Edit: I think instead of idxmax you are looking for something like

len(fileCSV.index)

edited Oct 26 '22 at 21:00

answered Oct 26 '22 at 20:30

bitflip

3,436
1
3
22

Can you expand meaning of `There is no such value for a series of url names (strings)` – xlmaster Oct 26 '22 at 20:33
2

Pandas calls numpy `.argmax` to do this function. `argmax` only works on numeric values. Pandas stores strings as objects, which numpy doesn't handle. – Tim Roberts Oct 26 '22 at 20:34
1

I assume you have a list of urls in your .csv file. Perhaps we can give a better answer if you show us an example of your data and your desired output. – bitflip Oct 26 '22 at 20:35
@bitflip I added visual of csv – xlmaster Oct 26 '22 at 20:53
What output do you want to get? The number of rows? (In this case 3) – bitflip Oct 26 '22 at 20:54
@TimRoberts if I could make `int(fileCSV['columnURLnames'])` sth like it, from columnB values? – xlmaster Oct 26 '22 at 20:54
@bitflip the index of max value in column B I will use in loop instead of `len(fileCSV.columns(1))` – xlmaster Oct 26 '22 at 20:57

reduction operation 'argmax' not allowed for this dtype

1 Answers1