How to create multiple Pandas time series from a data frame using python?

Question

I am starting with a Pandas data frame that looks like this:

   Type Date Number
1   A    x     y
2   B    x     y
3   A    x     y
4   B    x     y
5   A    x     y

I want to create separate time series for Type A data and Type B data separately. What is the most efficient way of doing this?

I am considering creating two different data frames from this where each data from only has data from one type and then converting each of the separate data frames to a series. However I don't know how to do this either.

Extended question: Is there a way to do this if you don't even know how many different types there are?

So far I tried checking to see if the type is of the type I want by using df["type"] == A, and this doesn't give me a full data frame back just a data frame saying if the type was true or false.

Additional information:

My goal is to create separate pandas time series using the date and number data for type A and type B separately.

I tried the following:

df.groupBy("Type").apply(lambda x: x.Date)

The above function works but only returns one column.

df.groupBy("Type").apply(lambda x: (x.Date, x.Number))

The above function doesn't work and returns something that is not what I want at all.

Expected Output:

Type    Date Number
 A   1   x     y
     3   x     y
     5   x     y
 B   2   x     y
     4   x     y

Please add an expected output.

Fabio Lamanna
– Fabio Lamanna

2015-10-17 08:36:31 +00:00
Commented Oct 17, 2015 at 8:36 — Fabio Lamanna
– Fabio Lamanna, Commented Oct 17, 2015 at 8:36

hellpanderr · Accepted Answer · 2015-10-17 11:54:33Z

1

If you want to group dates by type and put them into a separate Series, you can do the following.

Group by type: grouped = df.groupby('Type')

Get the date from every group: dates = grouped.apply(lambda x:x.Date)

dates now looks like this:

Type   
A     1    x
      3    x
      5    x
B     2    x
      4    x

You can access Series by type name: dates.A, dates.B etc

So far I tried checking to see if the type is of the type I want by using df["type"] == A, and this doesn't give me a full data frame back just a data frame saying if the type was true or false.

df["type"] == A gives you a boolean mask which you can plug back into a dataframe: df[df["type"] == A] But this is a very basic pandas operation, take a look at the official tutorial, there are a lot of examples: http://pandas.pydata.org/pandas-docs/stable/tutorials.html

answered Oct 17, 2015 at 11:54

hellpanderr

5,9563 gold badges42 silver badges50 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

pr338 Over a year ago

This is helpful! How do I get dates to have 3 columns including one for the x and one for the y?? I want to create a time series for the dates which are in x and the numbers which are in y.

pr338 Over a year ago

I tried this: grouped.apply(lambda x: [x.Date, x.Number]) and it didn't work.

hellpanderr Over a year ago

Can you post an example how should output look like?

pr338 Over a year ago

I posted some additional information above! Basically my end goal is to have 2 time series with date and the number information for each group separately. I'm not actually totally sure what a time series looks like.

pr338 Over a year ago

I added what I wanted the output to look like.

|

Collectives™ on Stack Overflow

How to create multiple Pandas time series from a data frame using python?

1 Answer 1

6 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

6 Comments

Your Answer

Sign up or log in

Post as a guest

Related