Sudden switch small Number to DateTime.

[wiluite]

Hello,

The csvkit version is 1.5.0

echo 13/12/2024, 5.5 | csvstat -H

1. "a"

   Type of data: Text
   Contains null values: False
   Non-null values: 1
   Unique values: 1
   Longest value: 10 characters
   Most common values: 13/12/2024 (1x)

2. "b"

   Type of data: Number
   Contains null values: False
   Non-null values: 1
   Unique values: 1
   Smallest value: 5.5
   Largest value: 5.5
   Sum: 5.5
   Mean: 5.5
   Median: 5.5
   Most decimal places: 1
   Most common values: 5.5 (1x)

Now I want to see Dates in the 1st column:

echo 13/12/2024, 5.5 | csvstat -H --date-format "%d/%m/%Y"

1. "a"

    Type of data:          Date
    Contains null values:  False
    Non-null values:       1
    Unique values:         1
    Smallest value:        2024-12-13
    Largest value:         2024-12-13
    Most common values:    2024-12-13 (1x)
   

2. "b"

   Type of data: DateTime
   Contains null values: False
   Non-null values: 1
   Unique values: 1
   Smallest value: 2025-05-05 00:00:00
   Largest value: 2025-05-05 00:00:00
   Most common values: 2025-05-05 00:00:00 (1x)

But now, I suddenly see a datetime in the 2nd column.
(If, however, I provide say 5.50 instead 5.5, or a larger number, than that's still number.)

Should I upgrade to the recent version?
Thanks.

[jpmckinney]

I recommend upgrading both csvkit and agate (a dependency) to their latest versions and retrying.

[wiluite]

Done:

wiluite@slax:~$ pip show agate
Name: agate
Version: 1.12.0
Summary: A data analysis library that is optimized for humans instead of machines.
Home-page: https://agate.readthedocs.org/
Author: Christopher Groskopf
Author-email: chrisgroskopf@gmail.com
License: MIT
Location: /usr/local/lib/python3.9/dist-packages
Requires: parsedatetime, Babel, leather, pytimeparse, python-slugify, isodate
Required-by: csvkit, agate-sql, agate-excel, agate-dbf

wiluite@slax:~$ pip show csvkit
Name: csvkit
Version: 2.0.1
Summary: A suite of command-line tools for working with CSV, the king of tabular file formats.
Home-page: https://github.com/wireservice/csvkit
Author: Christopher Groskopf and James McKinney
Author-email: chrisgroskopf@gmail.com
License: MIT
Location: /usr/local/lib/python3.9/dist-packages
Requires: agate-dbf, sqlalchemy, agate, importlib-metadata, xlrd, agate-excel, agate-sql, openpyxl
Required-by:

wiluite@slax:~$ echo 13/12/2024, 5.5 | csvstat -H --date-format "%d/%m/%Y"

1. "a"

   Type of data: Date
   Contains null values: False
   Non-null values: 1
   Unique values: 1
   Smallest value: 2024-12-13
   Largest value: 2024-12-13
   Most common values: 2024-12-13 (1x)

2. "b"

   Type of data: DateTime
   Contains null values: False
   Non-null values: 1
   Unique values: 1
   Smallest value: 2025-05-05 00:00:00
   Largest value: 2025-05-05 00:00:00
   Most common values: 2025-05-05 00:00:00 (1x)

Row count: 1