我有一个子集数据框架,并试图在"卷"列中的最大音量中找到索引。在这种情况下,它应该是索引1428,但是使用argmax或idxmcx,它给出了1431
combine1
Out[381]:
folder fn volume
1428 SF_20141231 IF1501_20141231.csv 162.0000
1429 SF_20141231 IF1502_20141231.csv 4.0000
1430 SF_20141231 IF1503_20141231.csv 6.0000
1431 SF_20141231 IF1506_20141231.csv 7.0000
1432 SF_20141231 TF1503_20141231.csv 4.0000
1433 SF_20141231 TF1506_20141231.csv 0.0000
1434 SF_20141231 TF1509_20141231.csv 0.0000
我将在哪里使用
combine1['volume'].idxmax(axis=0)
Out[385]: 1431
combine1['volume'].argmax()
Out[386]: 1431
这两个都是不正确的。如何解决此问题?
@user9240544,您需要使用to_numeric将列转换为float。请参阅下面的模型:
您注意到如果删除了combine1['volume'] = pd.to_numeric(combine1['volume'])
,则会将"音量"作为字符串获得,这就是您获得的结果。
raw_data = {
'folder': ['SF_20141231','SF_20141231','SF_20141231','SF_20141231','SF_20141231','SF_20141231','SF_20141231'],
'fn': ['IF1501_20141231.csv','IF1502_20141231.csv','IF1503_20141231.csv','IF1506_20141231.csv','TF1503_20141231.csv','TF1506_20141231.csv','TF1509_20141231.csv'],
'volume': ['162.0000','4.0000','6.0000','7.0000','4.0000','0.0000','0.0000']}
combine1 = pd.DataFrame(raw_data,index=[1428,1429,1430,1431,1432,1433,1434])
combine1['volume'] = pd.to_numeric(combine1['volume'])
combine1['volume'].idxmax(axis=0)
combine1['volume'].argmax()