TypeError when using pandas DataFrame.idxmax()

Question

My code (using pandas, pandas_datareader)

info = robinhood.RobinhoodHistoricalReader("AAPL", start=start, end=end, interval="5minute", span="day")
prices = pd.DataFrame(info.read(), columns=['close_price'])
minutes = prices.idxmax(axis=0) * 5

Gives a TypeError: reduction operation 'argmax' not allowed for this dtype

Any idea how to fix this? I just want to find the index of the max price in my dataset.

Could you give information about the column types? Try `prices.dtypes`. There is a good chance that a column is of string type while you expect float type. — Dric512, Apr 07 '18 at 20:11
Apparently they're type float, since dtypes says it's float64 — user9291465, Apr 07 '18 at 21:27
Is your first column (The dates) defined as the index? If not, it will try to find the maximum index for all the columns, including the date column. — Dric512, Apr 07 '18 at 21:46

score 1 · Answer 1 · answered May 21 '20 at 02:05

Try this:

info = robinhood.RobinhoodHistoricalReader("AAPL", start=start, end=end, interval="5minute", span="day")
prices = pd.DataFrame(info.read(), columns=['close_price'])
minutes = prices.astype(float).idxmax(axis=0) * 5

Assumes your row index is numeric. minutes is now a row showing the row_index*5 for the max cell value in each column.

TypeError when using pandas DataFrame.idxmax()

1 Answers1