Original subtitles

Welcome to the lecture on engineering.

Date and time variables.

Early on in the course we learned that date and time variables are those.

But it was that contained date time or a combination of both is something that we call a timestamp or

a date.

Timestamp we also saw some examples of date and date and time variables like for example date of birth

which in this case contains only dates.

Date of application for example for a loan which in this case contains the year and the month time of

accident.

For example which in this case contains only time for example as well payment date which contains in

this example a combination of date and time so all of these are examples of date and time variables

and in general we don't use them as such.

When we build machine learning mothers but we rather extract a variety of features from these characteristics

let's have a look on how we can do that.

So here we can see the anatomy of a daytime viable in this variable payment date that contains a date

and time we can see how we can easily derive the year of payment from this part of the variable.

We can also derive the months of the payment and with this variable we can create features that take

numbers from one to twelve indicating the month but we can also derive the quarter of the year in which

the payment was made or even the semester if we were interested similarly we can get they information

from the day part of the day time variable and we can get this information in the form of Monday Tuesday

Wednesday or instead numbers or we can create variables like is we can for example by looking at whether

these dates are Saturday or Sunday as we will see in the Jupiter notebook later on in the section we

can also look at the time part of the variable and derive our minutes and seconds.

Looking at the different components of the time and we can also get creative and look at whether this

payment took place in the morning or in the afternoon or in the evening more interesting.

We can calculate the time that passed between two variables.

For example we have the variable first payment date which contains a date on a time and then we have

another different variable that contains the information about the last payment date.

The similar format we can calculate the difference in time and we can express this in the number of

days that passed between the two payments or the number of months or the number of years or if the difference

is not as big as the one that I show in this example.

We can also calculate the difference in hours minutes and seconds.

Finally we need also to pay attention to the different time zones in which we can have two variables

particularly if our organization operates worldwide.

The Thompson empanadas is going to be indicated by a plus or a minus on the different of ours respect

to the center which is 0 so in order to operate with these variables we first need to put them in the

same star timestamp.

Whichever you want in this case I chose 0.

So as you can see 15 is two hours plus respect to zero.

So in the zero time zone this would be 1:00 o'clock whereas this one it's 3 p.m. but this time zone

is five times ahead of the center.

So in the center time it would be 10 a.m. and now that we have the values in the same time zone we can

operate and extract differences in time as we did in the previous light

in the rest of the section.

I'm going to cover two Jupiter notebooks in the first one I'm going to show you how to engineer dates

using banned and in the second one I'm going to show you how to engineer times using binders as well.

Feel free to read the notebooks in your own time or if you prefer.

Come with me in the next two videos and I will walk you through the notebooks.

See you in a bit.