Different standard deviation for each point Solved

Question

Hello.
I have a small problem. I would like to set different standard deviations for each bar (value belonging to the same series) of my histogram. I can't do it... When I enter a standard deviation, it gives me the same one everywhere but I can't set one for each bar of my histogram...

Could someone help me?

Raymond PENTIER · Answer

Do you know what a standard deviation is?
It's a concept applicable to a set of values, just like the maximum, minimum, or average.
Therefore, it makes no sense to have multiple standard deviations in a series of measurements.

More commonly referred to as standard deviation, it characterizes the width of the distribution. Mathematically, it is expressed as the square root of the variance, which measures the distribution of values around the center of the curve. Standard deviation (S) = Square root of the variance The standard deviation is a measure of dispersion ...

What you need is the margin of error for each measurement in the table, represented by an error bar in the graph; it can be proportional to the value or anything else. If you choose the standard deviation, the margin of error will be fixed, and the error bars will be of the same height.
https://www.cjoint.com/c/CEnc1iD3yiA

Best regards.

--
Retirement is great! Especially in the Caribbean ... :-)
☻ Raymond ♂

Dj1NNN · Answer

I'm not sure you understood... And I certainly didn't use the right term "series". I'm in biology, experiments must be reproducible. For that, we do triplicates. This means that for the SAME point, I have 3 different values, from which I can calculate a standard deviation. And this for each point of my experiment.
Let's say I measure a quantity at 10, 20, 30 minutes, I therefore have 3 values for each of my points. And my values 10, 20, 30 belong to the same series on my graph, but my standard deviation cannot be the same for the 3 values.
Basically, I did: STDEV(value 10 minutes sample A, " " " B, " " " C) =
Same for 20 and 30
So yes, my standard deviation will be the same for the 3 values measured at 10, 20, and 30 minutes, but what I want is to add a different standard deviation at 10, 20, and 30 minutes, since my standard deviations are not calculated from the same values.
It seems to me that the concept of standard deviation applies here, or am I confused...

Ps my histogram would therefore contain 3 bars: one at 10 minutes, one at 20, and one at 30.
The interest being to see if the standard deviations overlap, and whether the differences between my values at 10, 20, and 30 minutes are significant or not.

I hope this is clearer, maybe I'm confusing standard deviation with something else, but I would be surprised. It seems to me that we have always done it this way but oh well ^^

Raymond PENTIER · Answer

Hello.

I understood what you wrote, not what you forgot to specify ...
But I must be a little slow, because having 3 sets of measurements I would have been satisfied to display the three curves, which would clearly show the discrepancies for each measurement! But I'm not a biologist, and some analyses may escape me ...

https://www.cjoint.com/c/CEnsWq5RzDy

and see you later.

--
Retirement is nice! Especially in the Caribbean ... :-)
☻ Raymond ♂

ccm81 · Answer

Hello

It's hard to see what you're trying to achieve
A trial with error bars according to the standard deviations
https://www.cjoint.com/?3EnsVyisJEF

If I'm off track, could you please send the relevant part of your file in Excel 2003 format to cjoint.com and include the link in your next message?

Dj1NNN · Answer

Raymond, the goal is to present the data in the best way possible. 3 curves are unnecessary. The reader will wonder why there are 3 curves when I could have provided an average and standard deviations. It's less cluttered, clearer.

ccm81: it's the same kind of graph, 3 groups with 3 values in each group. However, your standard deviations don't all seem to be different.

Here’s an example

http://cjoint.com/?CEntk4BmYgI

Raymond PENTIER · Answer

Well!
It still took you 3 messages to completely explain your needs...
Is this "this graph" that you wanted, in fact?
If you disagree with the fact that I have distributed what you call standard deviation on either side of the mean value, you can always use columns K, N, Q instead of L, O, R.

Thank you for giving me the chance to have a little fun; it had been a while since I worked on error bars!
Retirement is nice! Especially in the Caribbean... :-)
☻ Raymond ♂

Dj1NNN · Answer

I asked for something simple: how to set DIFFERENT gaps on histogram bars, and you wanted to teach me what a standard deviation was.
It's not my fault if instead of answering the clear question asked, you try to read between the lines :)

And it's not me who calls it a standard deviation, it's indeed standard deviations... I checked with the researchers in the lab.
But yes, that's exactly the type of graph I wanted to create.

But it still doesn't tell me how you did it.

Different standard deviation for each point

7 réponses

Saracroche: the effective french app against phone spam

How to get the character µ on the android keyboard.

Texture issues in enshrouded

Access issue to my "contact me" profile

Windows spotlight is not working.

Black screen digital camera

Connecting rj45 cable with fiber

Sony bravia tv red led blinks 4 times

Free: associating with the freebox server in progress

Inuyasha in french