For a stripchart/jitter plot, is showing the median enough, or do you want to see more summary information about the shape of distribution embedded in the figure?
I think it also depends on the sample size, as you started to discuss. If I try to look at data from a giant drug trial, too many points=brain jitters. The current data set is at the upper size limit and I have mild jitters. My mental habit of methodically looking at axes/scale, labels, then distribution also hits bumps without the distribution explicitly plotted. I hit the same bumps on your Z-shape because when I went to find the axis labels I couldn't. Retraining my brain is slow, but worth it. Thank you for putting all these videos together. I am learning a ton and able to do a better job looking at my own data. You've also got some great resources to incorporate into teaching literature evaluation, which I appreciate.
I really like these plots and have seen in publications are fusion of these and bar plots which are quite good for those who really like the bars! Have you tried ggbeeswarm? I like that too as it does something similar but shows the density of points at each level on the y axis in a similar way to a violin plot.
Hi drjzee - Thanks for sharing ggbeeswarm, I'll look into it more. Some of the examples in their documentation look a lot like the output from geom_dotplot. I'll talk more about that in Wednesday's video with violin plots
@@Riffomonas Sir, It's related to your previous video....it also helps others and that paper also have correlation plots and I think you haven't made any video on that correlation plot also. Please Sir check it out once.
Hey Adolfo - thanks for the question! If you look back two episodes earlier, I went through this. Here is the link to the timestamp where I generate the count and add it to the x-axis label ruclips.net/video/IPey9jk_NOQ/видео.html
Thanks Cyrille! Stay tuned for the next episode where I will talks about boxplots. My concerns with boxplots is no one seems to know what the whiskers represent. Also, if you only have a few points, the box seems to be trying to say too much. For this data set, I think the boxes, without whiskers might work well. Keep watching 😊
For a stripchart/jitter plot, is showing the median enough, or do you want to see more summary information about the shape of distribution embedded in the figure?
I think it also depends on the sample size, as you started to discuss. If I try to look at data from a giant drug trial, too many points=brain jitters. The current data set is at the upper size limit and I have mild jitters. My mental habit of methodically looking at axes/scale, labels, then distribution also hits bumps without the distribution explicitly plotted. I hit the same bumps on your Z-shape because when I went to find the axis labels I couldn't. Retraining my brain is slow, but worth it. Thank you for putting all these videos together. I am learning a ton and able to do a better job looking at my own data. You've also got some great resources to incorporate into teaching literature evaluation, which I appreciate.
@@annazemke493 Thanks Anna - glad you're enjoying the videos!
Thank you so much! I was lost in the code for a homework assignment. My jitter plot looks leagues better now.
Fantastic! Glad to be of help 🤓
'horizontal variation within each category is determined randomly to separate the points'
-wise man
Thank you so much for your explanation!!
Thank you my great Master Shifu.
Ha! Thanks for watching
I really like these plots and have seen in publications are fusion of these and bar plots which are quite good for those who really like the bars! Have you tried ggbeeswarm? I like that too as it does something similar but shows the density of points at each level on the y axis in a similar way to a violin plot.
Hi drjzee - Thanks for sharing ggbeeswarm, I'll look into it more. Some of the examples in their documentation look a lot like the output from geom_dotplot. I'll talk more about that in Wednesday's video with violin plots
your videos are so useful, lots of thanks.
How do you comment (#) more thant one line at the same time?
I highlight the lines and then use the shortcut to comment them. On my mac it's shift-command-c
Sir, I have Mailed You a Paper In which I am confused that they Have used Slope Plot or Simple line Plot to Show F/B ratio Throughout childhood
Sorrry but I don’t have the bandwidth to help with individual issues
@@Riffomonas Sir, It's related to your previous video....it also helps others and that paper also have correlation plots and I think you haven't made any video on that correlation plot also. Please Sir check it out once.
Hello. How do you put N=155 under the labels (axis "X")
Hey Adolfo - thanks for the question! If you look back two episodes earlier, I went through this. Here is the link to the timestamp where I generate the count and add it to the x-axis label ruclips.net/video/IPey9jk_NOQ/видео.html
I confirm I disagree with you I greatly prefer boxplot, but as always is quite interesting to watch your video.
Thanks Cyrille! Stay tuned for the next episode where I will talks about boxplots. My concerns with boxplots is no one seems to know what the whiskers represent. Also, if you only have a few points, the box seems to be trying to say too much. For this data set, I think the boxes, without whiskers might work well. Keep watching 😊
@@Riffomonas I'm looking forward....
For me, it seems easier to explain whiskers rather than jitter plot, but I see your point ....
not working in my code
Are you using element_markdown?
@@Riffomonas No I am using only element_text in axis.text.x to make bold the words
@@rishikeshdash12 element_mardown is what you need instead of element_text
@@Riffomonas thank You so much Sir
@@Riffomonas Thank you sir...I did it.
how to add error bars here?
See the episodes on stat_summary
@@Riffomonas Thank you sir
how to increase jitter size?
jitter.width
@@Riffomonas inside the geom_jitter() function?
@@Riffomonas Sir it's not working inside the geom_jitter function