CODEWITHSUNDEEP
google-bigquerybigquery-dataframes
Chetana Tanmayee
Chetana Tanmayee

Reputation: 1

How to add days to a timestamp column in bigquery dataframes

I have a bigquery dataframe which has a timestamp column as shown below. I want to create a new column by adding 7 days to the creation_ts column.

creation_ts column

I have tried it as follows

`**df["TEMP_LABEL_END_TS"] = df["creation_ts"] + timedelta(days=7)**`

I am getting the following error message . ValueError: Literal did not coerce to a supported data type: interval('D').

Can anyone help me with this.

Upvotes: 0

Views: 72

Answers (1)

Tim Swena
Tim Swena

Reputation: 14786

Pandas-like timedelta behavior is not yet supported. I have filed a feature request on your behalf at https://github.com/googleapis/python-bigquery-dataframes/issues/719 and assigned it to a BigQuery DataFrames (bigframes) teammate who has most recently worked on datetime support.

In the meantime, you can workaround by casting to INT64 and back.

import bigframes.pandas as bpd

bpd.options.bigquery.project = "swast-scratch"

df = bpd.read_gbq_table("my-proj.my_dataset.my_table")
bpd.to_datetime(
    df['timestamp_col'].astype('Int64') + (7 * 24 * 60 * 60 * 1_000_000),
    utc=True,
    unit="us",
)

Upvotes: 0

Related Questions