+ Reply to Thread
Results 1 to 8 of 8

Combining / superimposing data series with different date ranges

  1. #1
    Registered User
    Join Date
    03-26-2018
    Location
    Wiltshire, UK
    MS-Off Ver
    365
    Posts
    4

    Combining / superimposing data series with different date ranges

    Hi, first post, so please be gentle. I've done a search but not found what I'm looking for, exactly.

    I have some performance data, collected via two different methods.

    One was Windows performance counters, gathered every 5 seconds over a period from, let's say 07:00 to 12:30.
    The other is some database stats, gathered every 5 MINUTES over a rather shorter period, with one or two gaps, let's say from 07:20 to 11:45

    I can chart them both separately, but I wonder if it's possible to combine them, and have Excel plot the data points in their correct relative position on the X axis? How would I achieve this?

    Many thanks

    Nick

  2. #2
    Forum Guru
    Join Date
    04-13-2005
    Location
    North America
    MS-Off Ver
    2002/XP and 2007
    Posts
    15,844

    Re: Combining / superimposing data series with different date ranges

    Not much to go on. It would be helpful to upload a sample workbook with some sample data and the charts you are currently using. Then we can see better about how to combine the two charts into one.

    My first thought is that this will be fairly easy with a scatter chart. Each data series can have its own X-values range, and Excel should easily plot the points where they should be. You don't say what chart type you are currently using. Would a scatter chart be suitable?
    Quote Originally Posted by shg
    Mathematics is the native language of the natural world. Just trying to become literate.

  3. #3
    Registered User
    Join Date
    03-26-2018
    Location
    Wiltshire, UK
    MS-Off Ver
    365
    Posts
    4

    Re: Combining / superimposing data series with different date ranges

    OK, here's the workbook with a little bit of anonymisation. You can see there are 2 sets of stats in the first two sheets, and a sheet with some line charts on. Currently I'm just plotting 2 or 3 of the disk stats, and the other stuff in a separate chart. However the timelines are broadly similar (although with different intervals as explained above, so it'd be nice to combine them.

    A scatter chart could be used if there's no other option, but there are enough different metrics that it's going to get confusing very quickly.
    Attached Files Attached Files

  4. #4
    Forum Guru
    Join Date
    04-13-2005
    Location
    North America
    MS-Off Ver
    2002/XP and 2007
    Posts
    15,844

    Re: Combining / superimposing data series with different date ranges

    I see two options, and I am not sure which will be more "confusing" for you:

    1) Use a scatter chart and assign the appropriate x value range for each series. I'm not sure what is confusing about this. I expect it will be tedious to go through each data series assigning x value ranges, but it should be easy.
    2) Combine all tables (your sample only has two tables, are there more?) into a single table. In this table the x values column will combine and/or merge the time stamps from the source tables. This table will also have a lot of "blank" entries where a given data series does not have an entry corresponding to a given time stamp. Then you can build your scatter chart (I would still use a scatter chart so that the times are treated as times and not evenly spaced text) from this table, using the same range for the x values of all series. You may want to review this (https://peltiertech.com/mind-the-gap...g-empty-cells/ ) to understand how Excel will treat all those blank cells -- especially if you end up using formulas to build this merged table. This approach avoids the tedium associated with editing each data series, but it adds the tedium of building another table.
    2a) Get and Transform might be able to easily merge these tables, but I am not familiar with this utility.

    In either case, I would use a scatter plot so that the times are real time values and not just text categories, just in case it makes a difference. I am not sure which will be more confusing.

    You have a lot of data in those tables. Trying to put too much information into a single chart will cause confusion whichever way you choose to do this. If your use of confusing alludes to getting too much data into a chart, you may want to consider what you are trying to show with these charts and figure out the minimum data to put into the chart that will communicate what you want.

  5. #5
    Registered User
    Join Date
    03-26-2018
    Location
    Wiltshire, UK
    MS-Off Ver
    365
    Posts
    4

    Re: Combining / superimposing data series with different date ranges

    Thanks.

    By confusing I was indeed alluding to the amount of data, although I was primarily referring to the number of series (i.e. the Y axis). I can reduce that, and maybe present different charts for different purposes. I can create overview and detail charts for different categories.

    The main issue I had was the discrepancy in the time intervals, but I get your idea about combining the two tables (no there aren't any more). I guess I just paste the data from one table on the end of the other and then sort it by time

    I wonder if there's a way of plotting the data so it samples the busy table at a bigger interval (say every 10 points instead of every 1). I'll do some more research... Or maybe just combine a selection of the records - that should be reasonably easy.

  6. #6
    Forum Guru
    Join Date
    04-13-2005
    Location
    North America
    MS-Off Ver
    2002/XP and 2007
    Posts
    15,844

    Re: Combining / superimposing data series with different date ranges

    There are probably a few different ways to get a sample of points from a large table. Something as simple as =INDEX(main_table,ROW(A1)*10-9,column#) would sample every ten rows, for example.

  7. #7
    Registered User
    Join Date
    03-26-2018
    Location
    Wiltshire, UK
    MS-Off Ver
    365
    Posts
    4

    Re: Combining / superimposing data series with different date ranges

    Many thanks again. I got a selection by a rather cruder method (data filter then a text filter on the seconds value), but your idea is much nicer.

    H.A.N.D.

  8. #8
    Registered User
    Join Date
    03-31-2019
    Location
    Montreal, Canada
    MS-Off Ver
    Professional Plus 2013
    Posts
    1

    Re: Combining / superimposing data series with different date ranges

    MrShorty-

    With your help in this thread and a bit of manipulation of data I solved my time series superimposing challenge.
    I needed a line graph so had to repeat the numbers in the blank spaces. Also had to merge the two series of dates from each of the data sets which went quicker with a macro. But apart from that, a snap.
    Many thanks.

+ Reply to Thread

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. Finding unique days from a series of date ranges
    By sharp4 in forum Excel Formulas & Functions
    Replies: 1
    Last Post: 01-30-2018, 01:47 PM
  2. Replies: 1
    Last Post: 10-25-2015, 08:51 AM
  3. Replies: 2
    Last Post: 05-12-2014, 10:24 PM
  4. Pivot chart: combining data series
    By Brython in forum Excel Charting & Pivots
    Replies: 3
    Last Post: 11-21-2013, 08:57 AM
  5. Combining two date ranges into one
    By macutna in forum Excel Programming / VBA / Macros
    Replies: 7
    Last Post: 07-27-2011, 11:26 AM
  6. combining two time series data
    By h1981 in forum Excel General
    Replies: 3
    Last Post: 06-25-2011, 04:23 AM
  7. Combining Cells for new Data Series
    By _PD_ in forum Excel General
    Replies: 3
    Last Post: 07-06-2010, 04:47 AM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Search Engine Friendly URLs by vBSEO 3.6.0 RC 1