Groupby Grouping Data

Quick reference

• 00:05 Okay, in this video, I want to talk about grouping by.
• 00:08 And group by allows us to just like a sounds, group different data together.
• 00:12 And a lot of times we're going to group by column and
• 00:15 then do something with that grouped data.
• 00:17 So I've just created a basic data frame here, and I created some data.
• 00:23 And we're going to have columns of corporation, employee, and salary.
• 00:27 And the columns in there are going to be corporations.
• 00:32 There's a list of the corporations we're going to have.
• 00:34 For the employees, there's a list of employees.
• 00:36 And salaries, there's a list of their salaries.
• 00:38 Assume this is like in thousands, right?
• 00:41 So, you know, 100,000, I put 180, 80,000, 80, etc.,
• 00:45 just to make it easier to look at.
• 00:46 So then we can run this and just create a data frame like we've done before and
• 00:51 pass in that data.
• 00:52 And if we look at this, we can just see we've got Corporation, Employee and
• 00:57 Salary, we got some Amazon, Tesla and Apple, some employees and their salaries.
• 01:02 So let's say we want to group by the company, right?
• 01:05 We can go df.groupby, and then just call whatever column,
• 01:11 so we got the corporation.
• 01:13 Now, if we run this, we get this DataFrameGroupBy object, and
• 01:17 this is just a location in memory.
• 01:20 So to actually do something with this, let's add this to a variable.
• 01:26 Now, we can also just call this, and we'll get that same object, but
• 01:31 now we can do stuff to this variable.
• 01:33 So say, for instance, we want to find the sum of all the salaries for
• 01:38 each company, we could call the sum function.
• 01:42 Remember, we looked at these functions way back at the beginning of the course.
• 01:45 So the sum of Apple salaries is 180.
• 01:48 So we can confirm that 100 plus 80 is 180.
• 01:50 Apple's 246, Tesla's 330.
• 01:53 We could find the average, the mean, right?
• 01:59 90, 123, and 165, very cool.
• 02:02 Now notice, it's not putting out anything for employees.
• 02:05 And that's because our Employee column, these are strings, right,
• 02:08 this is text, and Python can't find the average of text, right?
• 02:12 It needs numbers.
• 02:12 But interesting, we can call the max function, remember the min and the max?
• 02:17 And we do get Employee column spit out for that.
• 02:20 And that's because Python can take strings and
• 02:24 find the max alphabetical value, so the highest alphabetical value.
• 02:29 So for instance, Steve, S comes later in the alphabet than J, John.
• 02:34 So the max would be Steve, kind of interesting, right?
• 02:36 We could do the same thing for min and we get that.
• 02:40 All kinds of things we could do.
• 02:42 We could do standard deviation, right,
• 02:44 ff we wanted to find the standard deviation of our salaries or the variance.
• 02:47 You could do that.
• 02:51 All those universal functions we looked at we could count, right?
• 02:56 So there are two employees per company.
• 03:01 So it's 2 and there are two salaries per company, so that's 2.
• 03:05 Very cool so I'm going to put this back to mean.
• 03:07 So that's how to group by.
• 03:09 There's one more I want to show you the describe.
• 03:11 This is kind of a good one because describe,
• 03:14 it spits out all that information in one big thing.
• 03:18 So we can find the count, the mean, the standard deviation, the min.
• 03:22 It looks like quartiles and the max.
• 03:26 So very, very cool.
• 03:27 And just a fun thing to spit,
• 03:29 out just to get a description of all your data in there.
• 03:32 So that's group by.
• 03:35 Pretty simple, but pretty powerful, and
• 03:38 you're going to use things like this a lot.
• 03:40 So in the next video, we'll look at merging, joining, and concatenating.

Lesson notes are only available for subscribers.