CODEWITHSUNDEEP
pythonpandassortingdataframe
MichaelRSF
MichaelRSF

Reputation: 896

max sort values not working for dataframe Python

I have a dataframe called stock1 consisting of 505 stocks, and their percentage change for a total of 52 weekly rows for each column. I wanted to find the stocks with the maximum value for the final weekly row, so I did:

 import pandas as pd

 st1 = pd.DataFrame({"FIN_MAX": stock1.tail(1).max().sort_values(ascending=False)})

The above code originally worked, but now it throws an error.

 TypeError: '>' not supported between instances of 'float' and 'str'

The closest I can get to is if I drop the sort_values(ascending=False), but then the values are not arranged in order from highest to lowest.

 st1 = pd.DataFrame({"FIN_MAX":stock1.tail(1).max()})

I've also tried the following code which also gives the same error as above.

 st1.sort_values(["FIN_MAX"], ascending=False)

I'm using Python 3.6.1 on Jupyter Notebook. Any help or alternatives would greatly appreciated.

It seems the error persisted due to a column created called Unnamed: 0 when I concatenated all the stocks from individual csv files. Once I did del stock1["Unnamed: 0"], it seems to have resolved the issue.

Upvotes: 1

Views: 362

Answers (1)

piRSquared
piRSquared

Reputation: 294218

IIUC:

To get the stock with the largest value

stock1.iloc[-1].idxmax()

To get the stock with the largest value and the value

stock1.iloc[-1].nlargest(1)

To get the n stocks with the largest values and their values

n = 5
stock1.iloc[-1].nlargest(n)

Upvotes: 1

Related Questions