chungus
Discoverer
do i need to know the basics of computer science before getting in to programming languages like SAS, R, PYTHON etc?
- Thread starter chungus
- Start date
Supt
Explorer
JavaScript is good if you want to learn the basics and then move to other languages like C, C++ or Rust the syntax is similar.
Python is easier to learn and a lot of Data Analysts use python, so you won't need to learn a new language.
While python doesn't use {} or ; and stuff, if you mess up the indentation/spacing your code won't work even if you have the right code.
For your use case, python is definitely the better choice.
Use VS Code with python extensions and a formatter.
And here's a channel I would recommend, Bro Code.
Python is easier to learn and a lot of Data Analysts use python, so you won't need to learn a new language.
While python doesn't use {} or ; and stuff, if you mess up the indentation/spacing your code won't work even if you have the right code.
For your use case, python is definitely the better choice.
Use VS Code with python extensions and a formatter.
And here's a channel I would recommend, Bro Code.
You can do both side by side (and what any traditional cs degree puts you through too). I suggest if you like courses check harvard's cs 50 (will teach python) or if you like books "Automate the Boring Stuff with Python". Great intro points into programming.
A common recommendation would be to start with python. But I think from a programmer perspective it is better to start with C or C++. They actually help you on developing your code writing skills rather than applying imported functions.
Also arduino is a good place to start embedded system programming where you could visually see the effects. Just turning on different LEDs connected to the board would help you understand what variables are and how to assign values to them and how they affect the program.
Running a motor with PWM or creating a pattern using LEDs helps in understanding the conditional statements.
Likewise there are many programming logic which you can practically apply and enjoy the results.
Paul McWhorter Videos on YouTube is a great place to learn arduino.
Also arduino is a good place to start embedded system programming where you could visually see the effects. Just turning on different LEDs connected to the board would help you understand what variables are and how to assign values to them and how they affect the program.
Running a motor with PWM or creating a pattern using LEDs helps in understanding the conditional statements.
Likewise there are many programming logic which you can practically apply and enjoy the results.
Paul McWhorter Videos on YouTube is a great place to learn arduino.
maybe a bit too late, but what gives – here for posterity
try not to dive into comprehensive tutorials yet. the less overwhelming it feels at first, the more likely you’ll stay invested in it long term. what really works is developing an organic curiosity about what you’re learning, and that this curiosity is rewarded with appropriate discoveries. focus on getting comfortable with the basics and intermediate concepts, and take your time with them – as slow as possible to start with. the more time you spend at this stage experimenting in inquiry and exploration, the better. a sense of exploration is what will keep you going and help you grow before you can really fasten the straps
@chungus It appears to be 6 months too late reply but hope this helps.
From your question it appears that you are dipping your hand in more than one pot. Let me break it down as I see it.
Data Analysis - These days anything you do with data they call it data analysis, and job title like data analyst and data scientist are given out like candy in a festival so not much use in searching for job titles to understand what that job does. But if you want to have a clear understanding of what it actually is then here is a brief rundown of it. Data analysis is about finding insight from available data, this data doesn't need to be in GB,TB range even a simple 5 row 1 column data can and will contain insight. Now someone would have seen this data and thought okay lets do these analysis(explained later, for now assume Analysis 1, Analysis 2, etc) and see what we get out of it. This is generally done by Data Scientist and to an extent Sr.Data Analyst also. If they feel that a new kind of analysis is to be undertaken then they will design/test/implement the algorithm also to do the same. Once that is done then this entire procedure will be passed on to Data Analyst who will do the actual analysis and interpret the results and write a detailed report that is technical in nature for the Data Scientist. The Data Scientist will in-turn write a business summary of that report for the management.
Tools of the Trade - To be a good data analyst you need to understand 3 things, Database, Analysis Tools, UI Design.
In the database section you should have clear understanding of database design, query language and database administration. Database design consist of data warehouse design, data warehouse scheme design, etc. Query language is SQL(on Oracle it is PL/SQL) for designing ETL(extraction, Transformation and Loading) procedures(there are how you convert production data into historical data for running analysis). and Database Administration will be about backup scheduling, resource allocation(CPU, RAM, and Diskspace), your company's DBA will have control but you will be interfacing with him to allocate resources for your process as you see fit.
In the Analysis Tools section, these are the tools that you use to analysis and run the test on the data. Here you can use R language, SAS, Microsoft BI, etc to do it. When I listed analysis tools I did not include any underlying models(or testing methodology) deliberately. Because these days most of the so called data analyst can be called by old title "script kiddies". The script to conduct any and almost all standard tests on data are available as scripts on both R, SAS(dont know about Microsoft BI), you just supply it with data and it will spit out the result. You can just print it out and call it done, most of these new age data analyst are these kind. What sets the real data analyst from the script kiddies are the real understanding of the statistical models by the data analyst. A good data analyst will not only be able to interpret the result but will be able to justify both the process and the data used for doing so. For example you can run a co-relation analysis on rise of sea level and the use of toilet water. Can a co-relation analysis be done on these two data sets? Yes, Does it makes sense to run the above mentioned test? No. A data analyst who has a good understanding of statistics will know why one data set can and not not be used for analysis. I choose a really bad example but I hope you got the idea.
In the UI Design section, Not all analysis results are printed out and given to executives, some of them are converted into KPI(Key Performance Indicator) variables and displayed to top level management in realtime to track issues/production/etc. So good data analyst should know, how to display a given data? what kind of format should the display be? how the layout of the display should be? among others. Because these information are/will be glanced over in seconds and within those couple of seconds the display should convey a lot of important information with absolute clarity. There is more but I hope you get the idea.
Now comes the hard part, you want to do data analysis and what to use SAS for your job. Now SAS is a package which can be used for doing data analysis, dashboard design, warehouse design, SAS Advanced Data Miner can do AI training too. So as you might be aware SAS is a all-in-one package. Said that you cannot get a job which uses SAS without somekind of certification from SAS. Now do you need to go to a tutor to pass SAS certification? NO a BIG NO. So when you register for a test they give a trial period to practise, do not waste even a single minute of it, and have a solid foundation in statistics before you approach that. With this you can easly pass the tests.
Can a person become a data analyst without knowing the above? yes but he will always be a low level cog and will never move up in position and the first to let go in a restructuring. So don't go that route.
I have a couple of questions if you dont mind answering,
1. What sector you are interested in? If it is financial sector(Banks, Mutual Funds, etc) then there is no better software package than SAS, as SAS has embedded itself in this sector very much.
2. Where are you located? Tied to the first question, If you target London based financial institutions with SAS certification and a basic financial certificate then you have greater chance of getting a higher paying job.
I have given a overview of what need to know w.r.t SAS, data analysis and computer science. If you want further help reply back and I will help you out as much as I can.
Regards
From your question it appears that you are dipping your hand in more than one pot. Let me break it down as I see it.
Data Analysis - These days anything you do with data they call it data analysis, and job title like data analyst and data scientist are given out like candy in a festival so not much use in searching for job titles to understand what that job does. But if you want to have a clear understanding of what it actually is then here is a brief rundown of it. Data analysis is about finding insight from available data, this data doesn't need to be in GB,TB range even a simple 5 row 1 column data can and will contain insight. Now someone would have seen this data and thought okay lets do these analysis(explained later, for now assume Analysis 1, Analysis 2, etc) and see what we get out of it. This is generally done by Data Scientist and to an extent Sr.Data Analyst also. If they feel that a new kind of analysis is to be undertaken then they will design/test/implement the algorithm also to do the same. Once that is done then this entire procedure will be passed on to Data Analyst who will do the actual analysis and interpret the results and write a detailed report that is technical in nature for the Data Scientist. The Data Scientist will in-turn write a business summary of that report for the management.
Tools of the Trade - To be a good data analyst you need to understand 3 things, Database, Analysis Tools, UI Design.
In the database section you should have clear understanding of database design, query language and database administration. Database design consist of data warehouse design, data warehouse scheme design, etc. Query language is SQL(on Oracle it is PL/SQL) for designing ETL(extraction, Transformation and Loading) procedures(there are how you convert production data into historical data for running analysis). and Database Administration will be about backup scheduling, resource allocation(CPU, RAM, and Diskspace), your company's DBA will have control but you will be interfacing with him to allocate resources for your process as you see fit.
In the Analysis Tools section, these are the tools that you use to analysis and run the test on the data. Here you can use R language, SAS, Microsoft BI, etc to do it. When I listed analysis tools I did not include any underlying models(or testing methodology) deliberately. Because these days most of the so called data analyst can be called by old title "script kiddies". The script to conduct any and almost all standard tests on data are available as scripts on both R, SAS(dont know about Microsoft BI), you just supply it with data and it will spit out the result. You can just print it out and call it done, most of these new age data analyst are these kind. What sets the real data analyst from the script kiddies are the real understanding of the statistical models by the data analyst. A good data analyst will not only be able to interpret the result but will be able to justify both the process and the data used for doing so. For example you can run a co-relation analysis on rise of sea level and the use of toilet water. Can a co-relation analysis be done on these two data sets? Yes, Does it makes sense to run the above mentioned test? No. A data analyst who has a good understanding of statistics will know why one data set can and not not be used for analysis. I choose a really bad example but I hope you got the idea.
In the UI Design section, Not all analysis results are printed out and given to executives, some of them are converted into KPI(Key Performance Indicator) variables and displayed to top level management in realtime to track issues/production/etc. So good data analyst should know, how to display a given data? what kind of format should the display be? how the layout of the display should be? among others. Because these information are/will be glanced over in seconds and within those couple of seconds the display should convey a lot of important information with absolute clarity. There is more but I hope you get the idea.
Now comes the hard part, you want to do data analysis and what to use SAS for your job. Now SAS is a package which can be used for doing data analysis, dashboard design, warehouse design, SAS Advanced Data Miner can do AI training too. So as you might be aware SAS is a all-in-one package. Said that you cannot get a job which uses SAS without somekind of certification from SAS. Now do you need to go to a tutor to pass SAS certification? NO a BIG NO. So when you register for a test they give a trial period to practise, do not waste even a single minute of it, and have a solid foundation in statistics before you approach that. With this you can easly pass the tests.
Can a person become a data analyst without knowing the above? yes but he will always be a low level cog and will never move up in position and the first to let go in a restructuring. So don't go that route.
I have a couple of questions if you dont mind answering,
1. What sector you are interested in? If it is financial sector(Banks, Mutual Funds, etc) then there is no better software package than SAS, as SAS has embedded itself in this sector very much.
2. Where are you located? Tied to the first question, If you target London based financial institutions with SAS certification and a basic financial certificate then you have greater chance of getting a higher paying job.
I have given a overview of what need to know w.r.t SAS, data analysis and computer science. If you want further help reply back and I will help you out as much as I can.
Regards
chungus
Discoverer
Given that you are from the pharmaceutical background and wish to pursue a carrier is clinical test data analysis, the I would highly recommend you start with a strong foundational knowledge in statistics. Without statistical knowledge you will never understand the tests nor its results. A regular text book on statistics will be difficult for you to follow so I suggest statistics books that are written for life science professionals. Those books contain almost all the tests and their result interpretations and also why those tests are conducted.
Secondly, As you are going with SAS, you need to learn to code the SAS, those syntax and process is not like any other so no matter what computer programming language you study it wont help. Said that the coding part of SAS is the easiest, so you can start with it and learn without previous exposure. Example of One-Way ANOVA. Knowing just the SAS and statistical methods will help you get a job but wont help you in carrier progression. In order to do that you need to understand the entire data analysis field(data storage, data analysis tools, data display, etc mentioned in previous post). It may sound difficult but if you take it one by one then you can do it. Python, R, SPSS and such wont help you, once you are in SAS you are struck with it, unless you are willing to retrain yourself. Though R is a great language, none of the research institutions use it as such and management want shiny graphs and realtime display, so use R to gain knowledge about statistical methods and for doing private analysis to gain working knowledge.
Hope this helps
chungus
Discoverer
So I can start directly with sas without learning any other statistical language that is easier first? I thought learning JavaScript might help but I gave up half way as it wasn't for me