120: Don’t Learn Python as a Data Analyst (Learn This Instead)

[00:00:00] [SPEAKER_01]: If you're trying to land your first data analyst job, I've got some sad news.

[00:00:04] [SPEAKER_01]: I don't think that you should learn Python.

[00:00:06] [SPEAKER_01]: Now, don't get me wrong, I love Python.

[00:00:09] [SPEAKER_01]: It's my favorite data tool and it's absolutely incredible.

[00:00:12] [SPEAKER_01]: It can basically do anything and it's an awesome tool to analyze data.

[00:00:16] [SPEAKER_01]: But there's just a few reasons I don't think you should learn it when you're just starting out.

[00:00:21] [SPEAKER_01]: In this episode, I'll explain why and tell you what to focus on instead.

[00:00:30] [SPEAKER_00]: Professionals Land Their Next Data Job

[00:00:32] [SPEAKER_00]: Here's your host, Avery Smith.

[00:00:35] [SPEAKER_01]: Python has really built momentum over the last 20 years and become one of the most used data tools

[00:00:41] [SPEAKER_01]: out there. I mean, I totally see why. It's free, it's open source, there's over 137,000 libraries

[00:00:48] [SPEAKER_01]: that include amazing data tools that make your life as a data analyst way easier like pandas

[00:00:53] [SPEAKER_01]: for data handling, plot leaf for data visualization, and scikit learn for machine learning.

[00:00:57] [SPEAKER_01]: So why don't I think that you should learn it?

[00:01:01] [SPEAKER_01]: Two words, Opportunity Cost.

[00:01:04] [SPEAKER_01]: Opportunity cost is the loss of potential gains from other alternatives when one alternative

[00:01:09] [SPEAKER_01]: is chosen. When you're first starting out in data, there are just better things to learn first.

[00:01:14] [SPEAKER_01]: Things that don't require near as much time as it does to learn Python.

[00:01:19] [SPEAKER_01]: You see, Python's learning curve looks something like this.

[00:01:22] [SPEAKER_01]: It takes a loooong time before you see any progress.

[00:01:26] [SPEAKER_01]: It's actually pretty maddening. In the beginning, I'm talking like the first six weeks,

[00:01:30] [SPEAKER_01]: you will probably feel that you're making little to no progress.

[00:01:33] [SPEAKER_01]: And there's a couple different reasons for this.

[00:01:35] [SPEAKER_01]: Number one, the installation is very tricky.

[00:01:38] [SPEAKER_01]: Python is actually just flat out hard to download and use.

[00:01:42] [SPEAKER_01]: Most people probably Google download Python, and that makes sense, that's what I would do too,

[00:01:46] [SPEAKER_01]: and click on the first link or the second link or the third link or the fourth link.

[00:01:50] [SPEAKER_01]: All of these direct you to python.org, which is a great website.

[00:01:55] [SPEAKER_01]: You can click it and download it and everything goes fine.

[00:01:57] [SPEAKER_01]: But most people really get stuck trying to figure out how to actually open it up.

[00:02:02] [SPEAKER_01]: The reason is downloading Python this way does not give you an IDE or interactive developer

[00:02:07] [SPEAKER_01]: environment or in simple terms, a good place to open up play around and actually code in Python.

[00:02:13] [SPEAKER_01]: You're really limited to using the command prompt when you download it this way,

[00:02:17] [SPEAKER_01]: which is very difficult for most users. I don't like the command prompt either.

[00:02:20] [SPEAKER_01]: And then once you have vanilla Python installed,

[00:02:22] [SPEAKER_01]: you'll eventually want to download and use one of those 137,000 libraries.

[00:02:27] [SPEAKER_01]: And that in itself is actually very difficult. It's hard to get these libraries on your machine.

[00:02:32] [SPEAKER_01]: So to be all these installation problems, I recommend using a browser only version to

[00:02:37] [SPEAKER_01]: get started something like deep note or hex, or you can download and install the anaconda

[00:02:41] [SPEAKER_01]: distribution as it comes with a lot of great IDEs and libraries and makes the whole process

[00:02:46] [SPEAKER_01]: a little bit more simple. The number two reason it's hard to get started is it's

[00:02:49] [SPEAKER_01]: programming language. It takes time to learn how to program. And as the term indicates,

[00:02:55] [SPEAKER_01]: it is a language. It wouldn't be easy to learn and say Japanese or Swedish in a week would it?

[00:03:00] [SPEAKER_01]: Neither is Python learning to program is a skill that takes time to learn.

[00:03:05] [SPEAKER_01]: There's even a whole new vocabulary you have to learn in the programming world.

[00:03:08] [SPEAKER_01]: Things like loops functions variables operators recursion list comprehension namespaces

[00:03:14] [SPEAKER_01]: just to name a few. These things are actually very complex concepts and it takes a while

[00:03:18] [SPEAKER_01]: to understand fluently what's going on. BI tools like power BI or Tableau have a less

[00:03:23] [SPEAKER_01]: steep learning curve because they're really just drag and drop or click and point very similar to

[00:03:28] [SPEAKER_01]: how PowerPoint works. It's just a lot easier to learn them faster. Personally, I think that you

[00:03:33] [SPEAKER_01]: can get to proficiency in one of these BI tools in about a month. But I think it would take

[00:03:38] [SPEAKER_01]: around six months to feel the same level of comfort in Python. So simply stated,

[00:03:43] [SPEAKER_01]: I personally think it's six times harder to start in Python than something like Tableau.

[00:03:48] [SPEAKER_01]: The third reason it's actually really hard to get started is Python can literally do anything.

[00:03:53] [SPEAKER_01]: You don't use SQL to make data visualizations as it's really just meant to query data. You don't

[00:03:57] [SPEAKER_01]: use tab low to build some sort of machine learning model as it's meant to create and

[00:04:02] [SPEAKER_01]: display dashboards. But Python, it's the Swiss army knife of data. It can literally do

[00:04:07] [SPEAKER_01]: anything that you can do in data and a lot more. So mastering Python would really require

[00:04:13] [SPEAKER_01]: you to master all forms of data analysis in one go, which is absolutely impossible. You can't do it.

[00:04:19] [SPEAKER_01]: But let's just say that you're not listening. You're like this dog. You ignored all of this

[00:04:23] [SPEAKER_01]: and you still think Python's easy or maybe you understand it's hard, but you still feel

[00:04:27] [SPEAKER_01]: the need to learn it. Well, let me give you this fact. Python is at best the third most

[00:04:32] [SPEAKER_01]: popular data tool for data analysts. This is according to Luke Bruce's job scraping data

[00:04:37] [SPEAKER_01]: and this is for all data analyst positions junior and senior. But from my personal

[00:04:42] [SPEAKER_01]: experience, I literally stare at data analyst jobs all day every day. My guess is that number

[00:04:47] [SPEAKER_01]: is actually much closer to 20% for entry level data analyst roles. Senior roles typically rely on

[00:04:53] [SPEAKER_01]: Python a lot more than these junior level roles do. That basically means four out of five jobs

[00:04:59] [SPEAKER_01]: don't even mention the use of Python. So think about it. Are you really going to slow

[00:05:03] [SPEAKER_01]: your job search down by six months to learn a tool that isn't even mentioned 80% of the time?

[00:05:09] [SPEAKER_01]: I'm no math expert, but the math isn't mathing here for me. That doesn't make a whole lot of

[00:05:14] [SPEAKER_01]: sense, especially when you consider my next point and that is that many companies will actually pay

[00:05:20] [SPEAKER_01]: you to learn Python landing that first data job getting your foot in the door in the data

[00:05:25] [SPEAKER_01]: world. That's the hardest part. You want to do that as soon as possible because once you're

[00:05:30] [SPEAKER_01]: in getting promoted and upskilling is so much easier. It is a lot easier to land a new job

[00:05:36] [SPEAKER_01]: when you already have a data job. A lot of the time you can actually use company time and resources

[00:05:41] [SPEAKER_01]: to learn new data skills. For example, when I worked at ExxonMobil, they wanted me to get

[00:05:47] [SPEAKER_01]: better at Python. So they actually let me study Python for one to two hours every week on company

[00:05:52] [SPEAKER_01]: time using the company's LinkedIn learning account. How awesome is that? So instead of

[00:05:57] [SPEAKER_01]: waiting six months to apply to jobs, I was able to get paid to learn. So I'll ask it this

[00:06:02] [SPEAKER_01]: way very simply, which would you prefer slowing your job hunt down to learn a skill that isn't

[00:06:07] [SPEAKER_01]: necessary 80% of the time or getting paid to learn that skill down the road? The choice is yours.

[00:06:13] [SPEAKER_01]: If you make the choice to learn Python, which once again is not what I would necessarily suggest,

[00:06:18] [SPEAKER_01]: here's what you should know and in what order. Number one variables, print statements,

[00:06:23] [SPEAKER_01]: three mathematical operations, four functions, five loops, six IDEs, seven libraries. The big ones

[00:06:33] [SPEAKER_01]: in data to know are pandas, matplotlibs, seaborne, plotly and numby. How to read data in with pandas,

[00:06:39] [SPEAKER_01]: descriptive analytics with pandas, filtering with pandas, and data visualization. I like

[00:06:45] [SPEAKER_01]: seaborne for getting started. I think it is the prettiest and the easiest to use. If you want

[00:06:49] [SPEAKER_01]: to learn more about what data skills you should be focusing on at the beginning, you can either

[00:06:53] [SPEAKER_01]: watch this video here or there'll be a link to it in the show notes down below. And if you still

[00:06:57] [SPEAKER_01]: want to learn Python, I'm excited for you and I love you anyways. If you love me, maybe you

[00:07:02] [SPEAKER_01]: can hit subscribe.