Results 1 to 6 of 6
  1. #1
    3 Star Lounger
    Join Date
    Feb 2004
    Location
    Kitchener, Ontario, Canada
    Posts
    228
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Expanding Chart (2002)

    I am look for some help. I have attached a file for the example. The last worksheet is a chart. What I want to do is make the chart so that the days where nothing is done can be removed from the chart but when something is added into the day it shows in the proper plac on the chart.

    Worksheets:
    Tracking Info - This is the header information and part tracking.
    Month Data Input - This is where the data for the chart is
    Trend Breakdown - This is the chart. there are more charts but I have removed them to make this as simple as possable.

    As you can see the chart is very empty with Days 1-11,14-23,27,29,31 not having any information at all (and they should not since no work was done on these days). What I would like to do is have the chart not show those days until their is a value in worksheet:Month Data Input under the "Total Accepted per period" cell. If this cell has a value then the day should show up on the chart. If the cell is blank then the day should not be on the chart.
    Attached Files Attached Files

  2. #2
    Plutonium Lounger
    Join Date
    Mar 2002
    Posts
    84,353
    Thanks
    0
    Thanked 29 Times in 29 Posts

    Re: Expanding Chart (2002)

    But the x-axis labels 1 ... 31 represent dates, so it would look very strange if there were missing values.

    If you really want to omit days without data from the chart, you should create a source table that only contains the dates that have data, and base the chart on that table.

  3. #3
    3 Star Lounger
    Join Date
    Feb 2004
    Location
    Kitchener, Ontario, Canada
    Posts
    228
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Expanding Chart (2002)

    The file is ment for any of our sort so each time you use the file on a new part the date will be different

  4. #4
    WS Lounge VIP sdckapr's Avatar
    Join Date
    Jul 2002
    Location
    Pittsburgh, Pennsylvania, USA
    Posts
    11,225
    Thanks
    14
    Thanked 342 Times in 335 Posts

    Re: Expanding Chart (2002)

    If you transpose your data table you can set up the same data range (only instead of a 2 rows it will be 2 columns) and add a datafilter onto the datatable. You can then filter the data so that only rows greater then zero will be displayed and the chart will only plot those points.

    Another option is to create an intermediate table (formulas or macros) which read the values into a table (in a new place or new sheet) and only has the non-zero entries to plot.

    But these option would seem to me to defeat the purpose of a "trendline" since you will be ignoring most of the data. I would think that a Defect chart (c Chart) would be better (keep all the points but add the average and the control limits.

    Steve

  5. #5
    3 Star Lounger
    Join Date
    Feb 2004
    Location
    Kitchener, Ontario, Canada
    Posts
    228
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Expanding Chart (2002)

    You must be Quality Background.....
    C chart works great if the days I am talking about are days with 0 rejects found. The problem comes from the fact that those days no parts were sorted at all. This creates false readings and people think no rejects were found. The chart should only show days where sorting ocured and no rejects found.

  6. #6
    WS Lounge VIP sdckapr's Avatar
    Join Date
    Jul 2002
    Location
    Pittsburgh, Pennsylvania, USA
    Posts
    11,225
    Thanks
    14
    Thanked 342 Times in 335 Posts

    Re: Expanding Chart (2002)

    <P ID="edit" class=small>(Edited by sdckapr on 10-Apr-08 07:41. Added PPS and attaqched modified file)</P>Yes I learned some Statistical Process and Quality Control for a previous assignment.

    Try this proportion chart. The calculations are at the bottom of the data (which is where the chart is derived from). As other days are added the control limits will tighten. As it stands with the current number of days (6) your UCL is so wide, you should not be surprised with up to 20% defectives (which is why most of these are built on larger numbers (30+) points.

    I would suggest not plotting monthly numbers if they are so sparse but plot them continually over the entire year (this would work better if the data were transposed so you could keep years of data instead of only 256...)

    Steve

    PS A way to plot the monthly numbers with tighter controls is to base the Average and the U/LCLs on some long-term data, not just this one month. The more data the tighter the control limit and the bettter estimate can be done for variation to try to remove and variation which is better to be ignored.

    PPS
    Here is a modified file. B52 has the overall proportion defects (total defects / total produced) and B51 has the number of days used to calculate it. These numbers now are calculated from the data at hand. If you enter historical proportion in B52 and the number of points used to derive it in B51 you can plot the current data vs historical limits.
    Attached Files Attached Files

Posting Permissions

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