Writing a text with scatter plot points - python-3.x

Ok, this is a bit of a funky idea, but is it possible to generate a dataset of points that, when plotted, display a desired text? I remember seeing this somewhere, but have no idea how to do it.

Related

Histogram in gnuplot - bins not separated one from another

Sorry for my English,
I would like to make a histogram in gnuplot, where bins are not separated one from another by vertical lines, something like it is shown in the following figure:
I usually use "plot [...] with boxes" command to make a histogram, but it leads to a histogram with such vertical lines, which make the histogram unclear for bigger number of bins in the plot.
My data consists of two values for every value of x.
My comment again as answer, according to the StackOverflow "rule": no answers in comments...
You are probably looking for:
plot ... with steps
Check help steps. You may also want to check this answer about steps.

Histogram using gnuplot

I'm trying to do a histogram in gnuplot, but I don't want my bins to have fixed width. I need the width of a bin to respect the actual width of the interval.
Also, I'm not really sure what to put in the data file or how to organize it.
Can someone help me with doing this?
I need the graph to look like
this

Show graph without label, possible?

I have a macro that produces graphs like this:
The data for the graphs looks like (100,0,0,94,0,132,0,126,115) and so on.
The dotted graph shows all the data (note the dots on the "floor", that's data with value 0). The lines connects the dots with values more than 0.
I have to do this using two graphs and it works but it's a bit messy at the bottom as two labels show for each color. The example above works ok but if you want to see more lines it gets really messy quick and hard to see. Is it possible to only show the dotted graph in the graph window but not the corresponding label? Basically I want one label per color at the bottom.
And yes, I know there's a graph with dots and lines but it doesn't work for me.
Since the data looks like (100,0,0,94,0,132,0,126,115) the lines looks like sawtooths going all the way back to 0 and then back up again. I want the lines to only show the data with value more than zero.
In your case, the simplest way would be to substitute all 0 values with =NA() function. Excel won't chart #N/A! values. You'll still see the category label in the axis, but Excel won't chart the actual zeros.

Set label font size from data column in gnuplot

I'm trying to plot using labels with varying font size. For example:
plot "some_data_file" using 1:2:(20-$3) using labels font sprintf("Helvetica,%d",variable)
Or something to that effect. The label's font size should be 20 minus the value in the third column. This line doesn't work, but I think it displays what I'm trying to do. Any ideas?
Look at cities.dem demo on gnuplot.info for a way to do this.
I'll steal the solution there, modified for you:
plot "some_data_file" using 2:3:(sprintf("{/Helvetica=%d %s}",\
20-$4, stringcolumn(4))) with labels

Is there a way to transform rectangle into any shapes in MS Excel?

In MS Excel, I applied conditional formatting to a cell A1 so that it's color changes from blue at value 0 and red at value 100. I copied that cell and pasted it as linked picture(I) specially.(Paste Special). The color of the rectangle picture now changes if value in original cell is changed. The problem is, the shape is only rectangular, it can be converted into square and diamond shape at best.If same rectangle can be transformed into different shapes(maybe by adding extra anchor on rectangle shape), a heat map can be created easily. Please share your insights about this thing if there is a way.
Following pictures may help understand the problem:
[Example][1]
If you want a heat map, please consider using a CHART not gazillions of picture objects, colored/placed/rotated/etc to ... mimic a chart.
First, google anything about "surface chart" in Excel, see how it looks in Excel, whatever. Just see it and think about it so you can compare it to your current approach.
You will observe some things, like:
it's 3D
it has just a few layers/colors
etc.
But really, all of them can be solved. Probably easier than your current approach.
you can easily turn off both axes and you can rotate it so the camera is totally straight top-down - then it looks flat as paper and noone can see it as 3D anymore
you can add more layers, you can set each of them to specific colors
etc.
Some resources:
multiple colors in surface chart
how to change rotation of 3d chart
geesh, I just found even a whole article/tutorial dedicated to creating heat map charts.
Please, read that last link and I'm pretty sure you will want to use that approach instead of doing picture puzzles.

Resources