[BUG-REPORT] Can't Convert a datetime column to a string column

[bsbalkar]

Description
I'm trying to convert a column containing data in datetime64 format to string.

Let's say I have a column named df['date'] containing dates in datetime64

I tried converting it to a string column using "to_string" but it ended up in the following error below:

df['date'].to_string()

TypeError: to_string(): incompatible function arguments. The following argument types are supported:
    1. (arg0: numpy.ndarray[float32]) -> vaex.superstrings.StringList64
    2. (arg0: numpy.ndarray[float64]) -> vaex.superstrings.StringList64
    3. (arg0: numpy.ndarray[int64]) -> vaex.superstrings.StringList64
    4. (arg0: numpy.ndarray[int32]) -> vaex.superstrings.StringList64
    5. (arg0: numpy.ndarray[int16]) -> vaex.superstrings.StringList64
    6. (arg0: numpy.ndarray[int8]) -> vaex.superstrings.StringList64
    7. (arg0: numpy.ndarray[uint64]) -> vaex.superstrings.StringList64
    8. (arg0: numpy.ndarray[uint32]) -> vaex.superstrings.StringList64
    9. (arg0: numpy.ndarray[uint16]) -> vaex.superstrings.StringList64
    10. (arg0: numpy.ndarray[uint8]) -> vaex.superstrings.StringList64
    11. (arg0: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    12. (arg0: numpy.ndarray[float32], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    13. (arg0: numpy.ndarray[float64], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    14. (arg0: numpy.ndarray[int64], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    15. (arg0: numpy.ndarray[int32], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    16. (arg0: numpy.ndarray[int16], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    17. (arg0: numpy.ndarray[int8], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    18. (arg0: numpy.ndarray[uint64], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    19. (arg0: numpy.ndarray[uint32], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    20. (arg0: numpy.ndarray[uint16], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    21. (arg0: numpy.ndarray[uint8], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64
    22. (arg0: numpy.ndarray[bool], arg1: numpy.ndarray[bool]) -> vaex.superstrings.StringList64

Invoked with: array([datetime.date(2020, 11, 4)], dtype=object)

Please help regarding this.

Software information

• Vaex version (import vaex; vaex.__version__):

• {'vaex': '4.4.0', 'vaex-core': '4.4.0', 'vaex-viz': '0.5.0', 'vaex-hdf5': '0.9.0', 'vaex-server': '0.6.0', 'vaex-astro': '0.8.3', 'vaex-jupyter': '0.6.0', 'vaex-ml':
  '0.13.0'}

• Vaex was installed via: pip in Pycharm and conda in jupyter notebook

• OS:
  Windows 10

Additional information
Please state any supplementary information or provide additional context for the problem (e.g. screenshots, data, etc..).

[JovanVeljanoski]

Hi,

there is a dedicated method for this in dt accessor. See this example:

import vaex
import numpy as np

date = np.array([np.datetime64('2009-10-12T03:31:00'),
                 np.datetime64('2016-02-11T10:17:34'),
                 np.datetime64('2015-11-12T11:34:22'),
                 np.datetime64('2003-03-03T00:33:15'),
                 np.datetime64('2014-07-23T15:08:05'),
                 np.datetime64('2011-01-01T07:02:01')], dtype='<M8[ns]')

df = vaex.from_arrays(date=date)

# Create formatted string
df.date.dt.strftime('%Y-%m-%d %H:%m:%s')

This way you can format your datetime columns into string depending on what you need.

I hope this helps

[bsbalkar]

Thanks @JovanVeljanoski

This worked for me.

But in earlier version of vaex one can convert datetime to text with df['date'].astype('str') or df['date'].to_string(). And it's more simple in this way. I hope you'll try to include it in newer versions of vaex.