Hi. I'm trying to set up a chart on a spreadsheet to best represent some data I've been asked to collect and I'm not sure if I'm doing this in the best way. Can anyone who has experience making charts help me determine whether I'm doing this the best way? Thanks.
The requirements:
We are developing software that performs a certain task.
We need to measure how long it takes to perform the task, and how that performance changes over time, as we receive new dev builds.
There are nine different methods of performing the task and all nine need to be measured and will sometimes be compared to each other.
The task involves some environmental variables so for each method we are going to perform the task five times and collect the lowest, highest and mean performance times for each method.
We will do this for each development build we receive. Because of the environmental variables, every time we collect data on a new development build we will perform the same tasks on a baseline shipped build, to control for outside changes.
The Excel chart that presents the data should show the minimum, maximum, and mean times for each method using a data point and error bar or something that looks like it. Crudely, like this:
_
|
|
|
|
O
|
|
_
I set up a scatter plot to list the times for all the methods using the baseline build, figured out how to get the error bars conform to the minimums and maximums, and figured out how to overlap the same data for the test build. But then I couldn't figure out how to offset the test build data so they weren't overlapping and one could see which error bar belonged to which. The Y axis is time and needs to be exact but the X axis is just a category and doesn't need to be exact, as long as it's clear which category each data set belongs to. But I haven't figured out how to accomplish this. Excel insists on treating each method as an exact number and not a group and I haven't figured out how to get around that.
I plan to use one template to create a new spreadsheet for each build, collect data and produce a chart on that spreadsheet for each build, and as each build is tested I will accumulate data from each spreadsheet in to a master that will show performance across builds, and performance of the baseline in parallel. I imagine the master will have to have several charts, probably broken up by method.
I can't post attachments so I can't show you what I've come up with or what I have in mind, I can only hope you get the idea.
First of all, is there a better way of doing what I'm trying to do? I've looked in to 3D scatter plots and Excel doesn't support them natively, plus it appears difficult to visually discern where in the 3D plot the data point is, even if it's illustrated with vertical bars.
And second, if this is the best way, how do I offset the different data sets so the test build and the baseline build don't overlap?
Thanks a lot. I know this sounds like a huge challenge, it has been to me, but I'm hoping it isn't so much for some of you.
The requirements:
We are developing software that performs a certain task.
We need to measure how long it takes to perform the task, and how that performance changes over time, as we receive new dev builds.
There are nine different methods of performing the task and all nine need to be measured and will sometimes be compared to each other.
The task involves some environmental variables so for each method we are going to perform the task five times and collect the lowest, highest and mean performance times for each method.
We will do this for each development build we receive. Because of the environmental variables, every time we collect data on a new development build we will perform the same tasks on a baseline shipped build, to control for outside changes.
The Excel chart that presents the data should show the minimum, maximum, and mean times for each method using a data point and error bar or something that looks like it. Crudely, like this:
_
|
|
|
|
O
|
|
_
I set up a scatter plot to list the times for all the methods using the baseline build, figured out how to get the error bars conform to the minimums and maximums, and figured out how to overlap the same data for the test build. But then I couldn't figure out how to offset the test build data so they weren't overlapping and one could see which error bar belonged to which. The Y axis is time and needs to be exact but the X axis is just a category and doesn't need to be exact, as long as it's clear which category each data set belongs to. But I haven't figured out how to accomplish this. Excel insists on treating each method as an exact number and not a group and I haven't figured out how to get around that.
I plan to use one template to create a new spreadsheet for each build, collect data and produce a chart on that spreadsheet for each build, and as each build is tested I will accumulate data from each spreadsheet in to a master that will show performance across builds, and performance of the baseline in parallel. I imagine the master will have to have several charts, probably broken up by method.
I can't post attachments so I can't show you what I've come up with or what I have in mind, I can only hope you get the idea.
First of all, is there a better way of doing what I'm trying to do? I've looked in to 3D scatter plots and Excel doesn't support them natively, plus it appears difficult to visually discern where in the 3D plot the data point is, even if it's illustrated with vertical bars.
And second, if this is the best way, how do I offset the different data sets so the test build and the baseline build don't overlap?
Thanks a lot. I know this sounds like a huge challenge, it has been to me, but I'm hoping it isn't so much for some of you.
