From cb03e938a7d7ae8c203a1dc7059a035d0a174a37 Mon Sep 17 00:00:00 2001
From: jcjveraa <3942301+jcjveraa@users.noreply.github.com>
Date: Wed, 23 Dec 2020 06:12:02 -0800
Subject: [PATCH] use Pandas timestamp instead of datetime timestamp

While the pandas dataframe seems to be in UTC (my inspector shows 'Z'  for zulu timezone, =UTC), Using datetime.datetime.timestamp seems to interpret the time as the local time. This results in different outputs depending on your local timezone (clearly visible in the ' Time of Day signal plot' ).

Using the pandas equivalent function works uniformly independent from local environment.
---
 site/en/tutorials/structured_data/time_series.ipynb | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/site/en/tutorials/structured_data/time_series.ipynb b/site/en/tutorials/structured_data/time_series.ipynb
index d06ebcfeaa9..1a08a63f1ff 100644
--- a/site/en/tutorials/structured_data/time_series.ipynb
+++ b/site/en/tutorials/structured_data/time_series.ipynb
@@ -398,7 +398,7 @@
       },
       "outputs": [],
       "source": [
-        "timestamp_s = date_time.map(datetime.datetime.timestamp)"
+        "timestamp_s = date_time.map(pd.Timestamp.timestamp)"
       ]
     },
     {