Closed
Description
This is not friendly (nor is this efficient in the slightest as it coerces all the data via .values). This needs to operate block-by-block (or column-by-column is prob fine).
from SO
In [11]: df = pd.DataFrame({'Date':['20/03/17 10:30:34','20/03/17 10:31:24','20/03/17 10:34:34'],
...: 'Value':[4,7,5]})
...:
...: df['Date'] = pd.to_datetime(df.Date)
...: df
...:
Out[11]:
Date Value
0 2017-03-20 10:30:34 4
1 2017-03-20 10:31:24 7
2 2017-03-20 10:34:34 5
In [13]: df.idxmax()
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-13-0c78e75af756> in <module>()
----> 1 df.idxmax()
/Users/jreback/pandas/pandas/core/frame.py in idxmax(self, axis, skipna)
5112 """
5113 axis = self._get_axis_number(axis)
-> 5114 indices = nanops.nanargmax(self.values, axis=axis, skipna=skipna)
5115 index = self._get_axis(axis)
5116 result = [index[i] if i >= 0 else NA for i in indices]
/Users/jreback/pandas/pandas/core/nanops.py in nanargmax(values, axis, skipna)
462 """
463 values, mask, dtype, _ = _get_values(values, skipna, fill_value_typ='-inf',
--> 464 isfinite=True)
465 result = values.argmax(axis)
466 result = _maybe_arg_null_out(result, axis, mask, skipna)
/Users/jreback/pandas/pandas/core/nanops.py in _get_values(values, skipna, fill_value, fill_value_typ, isfinite, copy)
181 values = _values_from_object(values)
182 if isfinite:
--> 183 mask = _isfinite(values)
184 else:
185 mask = isnull(values)
/Users/jreback/pandas/pandas/core/nanops.py in _isfinite(values)
224 is_integer_dtype(values) or is_bool_dtype(values)):
225 return ~np.isfinite(values)
--> 226 return ~np.isfinite(values.astype('float64'))
227
228
TypeError: float() argument must be a string or a number, not 'Timestamp'
fine on Series.
In [27]: df.Date.idxmax()
Out[27]: 2
In [28]: df.Value.idxmax()
Out[28]: 1