Welcome to the Lounge

NeverJustHere1-Jul-18 23:59

NeverJustHere

1-Jul-18 23:59

I'd use matplotlib and numpy/pandas in the python world.

numpy is a lower level high performance array structure.
pandas is a higher level data frame structure built on top of numpy

matplotlib is a graphing library that will take data frames and render a huge variety of visualisations.

Use an IDE like spyder and it's all quite easy and interactive.

And all free.

Chris C-B2-Jul-18 2:36

2-Jul-18 2:36

Thanks, but that's a bit like reinventing the wheel. I just have this bunch of data and need to display it in a graph outside the application, as a once-off. I was a bit gobsmacked that Excel couldn't do it. Somewhere there has to be something where you establish the X and Y axes of the graph from the max and min of the data, and then plot all the X,Y points from the dataset. The data collection points on the X axis are non-linear, but I need to display them along a linear X axis.

Do not escape reality : improve reality !

Rage2-Jul-18 0:46

Rage

2-Jul-18 0:46

Not free, but enough if it is a one-shot : imc FAMOS - data analysis framework - productive testing[^]

S Houghtelin2-Jul-18 1:24

2-Jul-18 1:24

I use the pirates favorite... R[^]

RStudio is nice IDE envirionment for R that makes it easier to use. You can import and export Excel and csv file easily. and there are plenty of support communities where you can find how-tos and examples.
RStudio – Open source and enterprise-ready professional software for R[^]

[Edit] - And it's open source.

It was broke, so I fixed it.

Chris C-B2-Jul-18 2:39

2-Jul-18 2:39

Thanks, but I really, realy, realy do not want to write one more line of code just to display the data just one time.

S Houghtelin2-Jul-18 2:51

2-Jul-18 2:51

So, what your really, realy, realy looking for is some CODZ PLZ URGNTZZ? Laugh | :laugh:

It was broke, so I fixed it.

Chris C-B2-Jul-18 3:10

2-Jul-18 3:10

Nah! I don't need no CODZ PLZ URGNTZZ. Don't really like the stuff. I am actually having poached salmon in a lemon sauce tonight. Laugh | :laugh:

S Houghtelin2-Jul-18 3:12

2-Jul-18 3:12

Is that with real lemons?

It was broke, so I fixed it.

Chris C-B2-Jul-18 3:29

2-Jul-18 3:29

Nah! Furniture polish - it's cheaper.

S Houghtelin2-Jul-18 3:33

2-Jul-18 3:33

Chris C-B wrote:
Nah! Furniture polish - it's cheaper.

plus if you're grilling, you get the added bonus of that flame thrower touch of excitement, the kids will love it!

It was broke, so I fixed it.

S Houghtelin3-Jul-18 2:37

3-Jul-18 2:37

So, how was the salmon? Wink | ;)

Anyway I got 'yer Codez. It's like 6 lines of code.

# Install & Load Library
if (!require("readxl")) install.packages("readxl")
library("readxl")

# Choose & Read Excel file, the 2 denotes the desired sheet number.
my_data <- read_excel(file.choose(), 2)

# Extract data from the desired columns. RStudio will provide drop-down with column names.
Ydata <- my_data$Column_With_Ydata
Xdata <- my_data$Column_With_Xdata

# Plot data
plot(Ydata, Xdata)

As with any code you can get as fancy as you want, but for a one time look see...

It was broke, so I fixed it.

Chris C-B3-Jul-18 19:18

3-Jul-18 19:18

The salmon was exquisite, thanks for asking. Served with fresh asparagus and pasta shells, and accompanied with ice-cold Löwenbräu. If my coding were as good as my cooking, I would be richer than Bill Gates! Laugh | :laugh:

As for the rest of it, thanks, but I have decided to write an add-on to Excel to fix this problem, as I seem to recall having suffered from it before.

Foothill2-Jul-18 3:13

Foothill

2-Jul-18 3:13

I just plugged two sets of random numbers into Excel (2013) and produced a proper scatter plot. I followed these steps to get the trend line: trendline example. Is your data set too big for Excel?

if (Object.DividedByZero == true) { Universe.Implode(); }

Chris C-B3-Jul-18 18:52

3-Jul-18 18:52

The scatter graph works fine for X-Y points, but the data collection points are equally spaced on the graph, where they are, in fact, non-linear to the point of being random. They are closely spaced at the start of the X-axis, becoming sparser further along - but they are essentially just points where data had been collected, and adhere to no mathematically defined function.

I want to impose a linear X-axis, say from 0 to 100, and have the graph points plotted along that linear axis. If I just use the X-Y data in a plot, then the X-axis is divided into equidistant points for the X value and the Y value plotted above it. If I add a third column of data to represent the X-axis, Excel accepts it but then draws two graphs, one for the X values and one for the Y values.

If it were not for the 11,000 odd data points, I would use a piece of graph paper...

Kirill Illenseer2-Jul-18 21:39

Kirill Illenseer

2-Jul-18 21:39

Scilab seems like the way to go. Several of my scientific co-workers use that for dealing with data sets doing all kinds of stuffs that physics does with data, including non-linear axes.

Chris C-B3-Jul-18 18:52

3-Jul-18 18:52

Thanks, I will check it out.

Member 99910752-Jul-18 21:53

Member 9991075

2-Jul-18 21:53

You could have a look at GNUplot, it certainly meets your requirement for free.

Chris C-B3-Jul-18 18:54

3-Jul-18 18:54

Having just Bingled that, it looks like it could be a winner - thanks.

Pete Kelley3-Jul-18 2:17

Pete Kelley

3-Jul-18 2:17

I don't have a solution to offer, but I'm wondering if you could give some usable example data?(rigged-up or otherwise) I certainly recognize that this simple need has arisen more times than I can remember. It's time to dabble in tool-building.
This kind of issue is a lot like when I only want to make a simple chicken-scratch sketch using old-school Paint but the system I happen to be using only has a major drawing program that takes too-long/forever to load up. Or I just want to type a sentence and a bloated word processor starts loading.
Great to have some of the suggestions - checking out GNUPlot and some of the free-or-inexpensive options.

Pete K.

Chris C-B3-Jul-18 18:58

3-Jul-18 18:58

Pete Kelley wrote:
This kind of issue is a lot like when I only want to make a simple chicken-scratch sketch using old-school Paint but the system I happen to be using only has a major drawing program that takes too-long/forever to load up. Or I just want to type a sentence and a bloated word processor starts loading.

'Twas always thus! In fact, although I said I didn't want to write a line of code, I am now thinking of writing an add-on for access to solve precisely this problem.

mngerhold3-Jul-18 3:10

mngerhold

3-Jul-18 3:10

At the risk of sounding dumb, I can't see the problem - what are 'non-linear X values' ? You can't just mean they are at unequal X-intervals, because surely that is trivial, even for Excel?

Chris C-B3-Jul-18 19:00

3-Jul-18 19:00

See my reply to Foothill above. If you know of how it can be done, I would love to hear about it.

mngerhold3-Jul-18 21:01

mngerhold

3-Jul-18 21:01

I must be being dumb. I read your reply above, could not visualise it. Although the chance of my being able to help is slim, could you post a scan of a scribbled hand plot to illustrate?

cmkrnl3-Jul-18 4:21

cmkrnl

3-Jul-18 4:21

I found ZedGraph maybe 10 years ago, and have driven it with C# for countless graphing needs since then.