laitimes

A boxplot with this parameter looks very different

A boxplot with this parameter looks very different

High value free SCI online drawing ImageGP

  1. The website architecture has been redesigned to quickly deploy more types of tools and richer forms of presentation of results.
  2. The drawing R code is open source for further customization after downloading.
  3. More parameters have been added, such as color modification, statistical markers
  4. Add more types of graphic drawing
  5. Add the function of online data conversion and merging, which is convenient for faster preparation of drawing data
  6. Add online analysis functions such as WGCNA, differential gene analysis, etc
  7. No registration is required to use it

This is an error reported by a user, and the error message is:

Error in cut.default(data[[xvariable]], xvariable_cut) : invalid number of intervals Calls: sp_boxplot -> cut -> cut.default Execution halted           

Through the user-provided reanalysis link, we can see that the parameter X-axis variable manual cut is set to 0=A 1=B 2=C 3=D 4=E, which is the source of the error.

What this parameter is used for, we'll talk about later. From this parameter setting, I guess that the user's intention is to convert numbers into letters for display. Let's take a look at the user's data, paste it into Excel, and use the data filter function (the data has been desensitized), and see that the grouped column is indeed 5 groups.

A boxplot with this parameter looks very different

The easiest way to do this is to write the correspondence between the number and the group name into an Exccel, as follows:

Num   Name
0   A
1   B
2   C
3   D
4   E           

Then paste the raw data from the above excel and the new number-group correspondence data into the Megre matrix function.

A boxplot with this parameter looks very different

Get the output matrix, and then input it into the Boxplot drawing tool, the other parameters are default, only modify the parameters highlighted in yellow in the figure, as follows:

A boxplot with this parameter looks very different

Once submitted, get the results

A boxplot with this parameter looks very different

It seems that the trend is obvious, add a statistical analysis to see if the difference is significant

A boxplot with this parameter looks very different

Let's look at the second way, what if it's based on cut? The parameters are modified as follows:

  • X-axis variable 选择 group
  • X-axis variable manual cut 输入 5,表示分成 5 组
A boxplot with this parameter looks very different

After submission, the result is obtained (the split interval is open and closed)

A boxplot with this parameter looks very different

Of course, the cut can also be set in this way, and the parameters are modified as follows:

  • X-axis variable 选择 group
  • X-axis variable manual cut 输入 -0.5,0.5,1.5,2.5,3.5,4.5,表示分成 5 组
A boxplot with this parameter looks very different

Once submitted, get the results

A boxplot with this parameter looks very different

There is a lack of a horizontal axis name renaming function, and the name of the horizontal axis can only be displayed according to the number division interval (this function involves requiring the user to input a lot of things, and it has not been added to the website for the time being, but you can download the code yourself and add this parameter locally).

The specific modifications are shown in the figure below:

A boxplot with this parameter looks very different