Charles, I agree 100% that a R course should start out with showing students what R can do rather than get lost in details in the beginning. See the forest first - then the trees.
Charles, In regards to the assignment operator I find that it makes sense if I think of it mathematically. For example, x = x + 5 makes no sense. However, x
Charles, My preference is to work and experiment with the data in a r script and then transition over to RMarkdown when I am closer to finished product. I find RMarkdown to be too slow as the go-to program.
I am working on setting up an R class for geoscience students in my department and thus looking around for good ideas on how to proceed with class format etc. My impulse is to start with data and show people what they can do with R (especially in the tidyverse). I am, however, running into some resistance from those who think Excel is just fine and are not interested in "learning how to code". Unfortunately R is not as popular in the physical sciences as it is in medicine, life sciences and social sciences. If you have tips/suggestions based on your long-term teaching then I would greatly appreciate hearing from you.
@@haraldurkarlsson1147 Hmm, that is interesting---R is fairly popular in the physical sciences where I've been, as are Python (interchangeable) or Fortran (for speed). I've never really had to "sell" it to people in physical/natural sciences as the utility of programming is usually immediately apparent. Getting Excel to do any complex tasks is a pain---and once you're doing anything complex in Excel you actually are programming already. Things like maps, automated data collection, high quality graphics, any real statistical estimation etc. are essentially impossible in Excel without getting deep into VBA programming, but they're common the sciences. If students are planning on producing research, they should be learning good research practices around reproducibility as well (necessary for publishing in many journals). Excel does not lend itself to this. Excel also has hard limits in the number of rows (a bit over a million) and columns (a bit over 16000) which are problematic if working with data of any real size---and it becomes ponderously slow at scale. Excel is a good tool for entering data or doing actual spreadsheet work, but it is a nonsensical choice for general data analysis unless working on incredibly basic tasks.
There's also the pragmatic approach to note that students in a quantitative field who are unable to do some sort of programming are drastically less employable. Inability or unwillingness to code would more or less rule out working in any serious quantitative scientific environment.
@@cclanfear Charles, That is interesting. I not sure what the deal is in my department. Most are happy with Excel or Minitab or MatLab and word for writing. I loath most of the Microsoft products (possibly because I am a Mac person). I actually used Word and Excel etc when they first came out and they worked well in those early days. These programs are just getting bigger and more cumbersome to operate. Nothing stay the same (like LaTeX does) and I am constantly fighting unexpected changes that occur in my powerpoint slides when the updates come out and mess up my presentations. Even though the faculty here are set in the old ways (even the 'younger' ones) I have a sense though that the students are more interested in R and realize that learning an open source program like R or Python (or even LaTeX) would enhance their employment chances. I point out to them that if we don't change with the times we will go academically extinct. You are in an area of the country where there is a lot of change and excitement about data "science" (not sure what that is actually) while I am in a more conservative portion of the country in more than just an academic sense. It is just difficult to get people to change. As someone said "Opinions don't change, people die". Thanks for the discussion.
Thanks a lot Charles, I'm start using R Studio. I am looking forward to learning from all of you day to day from Myanmar. Have a good day!
Thank you so much for uploading this class. It is one of the rarest courses specifically tailored for beginners.
Charles,
I agree 100% that a R course should start out with showing students what R can do rather than get lost in details in the beginning. See the forest first - then the trees.
This was a great lecture! Thanks so much for making these available. I look forward to following your lecture series on RUclips.
i truly enjoyed watching your lectures, almost like watching a movie for me! it's that kind of pleasure to watch your videos! lol
Thanks so much for making your videos public. I was struggling learning R all by myself.
Charles,
In regards to the assignment operator I find that it makes sense if I think of it mathematically. For example, x = x + 5 makes no sense. However, x
Hey, Great series with all R related content for data analytics in one playlist. Thanks
E
Literally everything is available on the website at clanfear.github.io/CSS508
@@cclanfear Hi Charles! I think there is a typo in the given link. Could you please re-check? Thank you :)
@@not_so_nabin Missing an S: clanfear.github.io/CSSS508
Charles,
My preference is to work and experiment with the data in a r script and then transition over to RMarkdown when I am closer to finished product. I find RMarkdown to be too slow as the go-to program.
Typically the same here, though more about doc length getting clunky quickly when writing articles.
I am working on setting up an R class for geoscience students in my department and thus looking around for good ideas on how to proceed with class format etc. My impulse is to start with data and show people what they can do with R (especially in the tidyverse). I am, however, running into some resistance from those who think Excel is just fine and are not interested in "learning how to code". Unfortunately R is not as popular in the physical sciences as it is in medicine, life sciences and social sciences. If you have tips/suggestions based on your long-term teaching then I would greatly appreciate hearing from you.
@@haraldurkarlsson1147 Hmm, that is interesting---R is fairly popular in the physical sciences where I've been, as are Python (interchangeable) or Fortran (for speed). I've never really had to "sell" it to people in physical/natural sciences as the utility of programming is usually immediately apparent. Getting Excel to do any complex tasks is a pain---and once you're doing anything complex in Excel you actually are programming already. Things like maps, automated data collection, high quality graphics, any real statistical estimation etc. are essentially impossible in Excel without getting deep into VBA programming, but they're common the sciences. If students are planning on producing research, they should be learning good research practices around reproducibility as well (necessary for publishing in many journals). Excel does not lend itself to this. Excel also has hard limits in the number of rows (a bit over a million) and columns (a bit over 16000) which are problematic if working with data of any real size---and it becomes ponderously slow at scale. Excel is a good tool for entering data or doing actual spreadsheet work, but it is a nonsensical choice for general data analysis unless working on incredibly basic tasks.
There's also the pragmatic approach to note that students in a quantitative field who are unable to do some sort of programming are drastically less employable. Inability or unwillingness to code would more or less rule out working in any serious quantitative scientific environment.
@@cclanfear
Charles,
That is interesting. I not sure what the deal is in my department. Most are happy with Excel or Minitab or MatLab and word for writing. I loath most of the Microsoft products (possibly because I am a Mac person). I actually used Word and Excel etc when they first came out and they worked well in those early days. These programs are just getting bigger and more cumbersome to operate. Nothing stay the same (like LaTeX does) and I am constantly fighting unexpected changes that occur in my powerpoint slides when the updates come out and mess up my presentations.
Even though the faculty here are set in the old ways (even the 'younger' ones) I have a sense though that the students are more interested in R and realize that learning an open source program like R or Python (or even LaTeX) would enhance their employment chances. I point out to them that if we don't change with the times we will go academically extinct.
You are in an area of the country where there is a lot of change and excitement about data "science" (not sure what that is actually) while I am in a more conservative portion of the country in more than just an academic sense. It is just difficult to get people to change. As someone said "Opinions don't change, people die".
Thanks for the discussion.
Hi, would it be poosible to post all the videos? Great lectures and I can use a ton of practice.
All the Spring 2021 videos are available in the playlist: ruclips.net/p/PLaHil0mtdmL4PdIxByyvKoE_-H4HqfmyM
Charles
Is "promise" what is referred to as "lazy" loading or evaluation - that is it will not take memory space until you run it?
thank u brian in neon ! I too am jack off all trades and Master of None -- contrary to popular vote i mean belief
Never mind... I am like a student not reading all the way (too lazy?).
:)