9 Replies Latest reply: May 16, 2016 2:23 PM by michal perry RSS

    week of year time series data

    michal perry


      I am new to datawatch Designer, appriciate your support.

      the date column in the data that I want to visualize is in a week of year basis (for example: 112016 which means week 11 of year 2016).

      is it possible to define this column as the time series column and display it as x axis in the line graph and other visualizations? I couldn't find any example for it in your documentation

      If not what is the best way to visualize it?



        • Re: week of year time series data

          Hi Michael Perry,


          We're not sure if this is possible in Designer right now as this was not included in the format for Time Series Data or there are no ways to customize the format to view Week/Year yet.


          We'll try to look for possible workaround for this but if there's none, we can create an enhancement request to add this feature in future release.


          Hope this helps.





            • Re: week of year time series data
              michal perry


              thanks for your reply, it indeed could be great to have it implemented.

              Also, wanted to note that the week in year is sometimes defined differently

              in different organisations. For example: I am Intel employee and our work

              week in year isn't always aligned with the calendar week.

              Anyway, my data is already in intel ww granularity (for example: 201601,

              201602, ...) and it could be great if i could use it as a time series data


              Regards, Michal.


              On Mon, May 9, 2016 at 7:45 PM, mperez <

            • Re: week of year time series data
              Theo Klemming

              Hi Michal


              To use the "true" timeseries visualizations, there must be a column in your data that can be understood as a complete date, like for example 2016-05-10 (May 10, 2016), optionally also including a time, like for example 2016-05-10 10:11:03.


              However, when you have YearWeek value like you describe, you can still create a line graph or a bar chart from your values across time. Assuming you have values like these, where both columns are numeric:


              (#) YearWeek(#) Value


              You would then duplicate the YearWeek column and convert the duplicate to Text:


              (#) YearWeek(#) Value(Abc) YearWeek1


              This will allow you to create for example:

              A Vertical Dot Plot with Value on height, and YearWeek1 on breakdown

              A Numeric Line Graph with YearWeek on x-axis, Value on y-axis, and YearWeek1 on Breakdown.


              It will create graphs like the ones showing below:




              Examples of other graphs that you can use in a similar way are the bar graph and the numeric needle graph.

              And, you can add your YearWeek column to a filter box to create a filter slider that allow you to focus in on a shorter time span.


              Best regards, Theo Klemming

                • Re: week of year time series data
                  michal perry

                  Hi Theo,


                  Thanks for your reply! This is exactly the approach I was thinking of and



                  Another visualization that I am looking for and I think is available only

                  for time series is the graph below. This one combines bar graph (for

                  volumes) with what you called in datawatch dot graph (for Yield rate-

                  number of good units)


                  Is there any option to have this kind of visualization?






                  thanks, Michal.


                  On Tue, May 10, 2016 at 11:23 AM, Theo Klemming <

                    • Re: week of year time series data
                      Theo Klemming

                      Hi Michal


                      For some reason, I can't see any picture in your previous post. However, what you describe sounds like a combination of graphs, and currently, we have that available only for timeseries data, in the timeseries combination chart, where you can overlay for example lines and dots. So, for that purpose, you may have to create a data set that matches each year-week to the last or first date of each week, as you prefer.


                      // Theo

                        • Re: week of year time series data
                          michal perry

                          Hi Theo,

                          got it, thanks!


                          I have another question which is regarding the navigation action and

                          passing parameters from a dashboard to other dashboard:

                          I have a dashboard (lets call it DB-2) that gets 3 parameters

                          (product_name, start_ww, end_ww).

                          I want to navigate from one dashboard (lets call it DB-1) to DB-2: for this

                          I created an action dropdown (single selection) and created a navigation

                          action that navigate to DB-2. The datasource for the dropdown action box is

                          a table which contains 3 colums: product_name, start_ww, end_ww. the

                          dropdown list is populated with the values from column 'product_name' of

                          this table. However, i want to pass the other 2 parameters as well:

                          start_ww and end_ww but i only sucess to pass the product_name. how can i

                          pass more parameters to DB-2?? do i need to use java script and script


                          thank you,




                          On Tue, May 10, 2016 at 2:55 PM, Theo Klemming <

                            • Re: week of year time series data
                              Edrun Yuen

                              HI Michal,


                              Kindly give us time to review your inquiry. if we find a way to achieve your expected outcome we will get back to you as soon as possible

                              • Re: week of year time series data
                                Theo Klemming

                                Hi Michal


                                An action drop down can only ever deliver a value for 1 single parameter. To deliver several parameter values in one go to another dashboard, you have 2 options (in none of the cases is script actions involved - the use of script actions is limited to communicating from a dashboard to other websites, systems,and applications - or to the web environment that embeds the dashboard).


                                The first option is easier to set up: Instead of using an Action Dropdown as the widget for setting your parameter values, use a visualization, and a data table that contains columns with values for each of your parameters, so in your case product_name, start_ww, end_ww. You will include only the product_name column in your Breakdown, creating one single visual item for each unique value of product_name. I recommend using the heat matrix, putting product_name on Rows. Then add the columns start_ww and end_ww to a filter box on the side. You can make the filter type Single Selection. When filtering on start_ww and end_ww, you will remove all rows that are not matching, and when launching your action from the visualization following that, you can configure your Navigation action to deliver values for all three parameters on DB-2. In the heat matrix (or other visualization), you can trigger the Navigation Action on double-click as long as there is only 1 level in the Breakdown+Rows+Columns.


                                The second option is more complex: use 3 Action Dropdowns, one for each parameter, and use 3 different data tables for delivering values to each Action Dropdown, where you start with start_ww, and then let the end_ww data table use the selected value of start_ww as the value of a parameter so that no end_ww valus displayed for selected are earlier than the selected start_ww. Lastly, the selected start_ww and end_ww can be used as values for parameters on the data table populating the Action Dropdown for product_name, thereby omitting any product_names that are not represented in the selected ww period. Now, to the most important part: In order to avoid having the user sent to DB-2 each time any one of the three Action Dropdowns are changed, you should manually add 3 intermediary parameters on DB-1 (right-click dasboard tab, select properties). Those parameters could be called for example im_product_name, im_start_ww, im_end_ww. Your Action Dropdowns must we wired to set the value of their im-twin parameter, and target DB-1 - not target DB-2 etc. Below the three dashboards, you will add a Button. The button will be configured to use new navigation action, which targets DB-2. The three parameters in use on DB-2 will get listed, and in the settings for the button, you will enter a parameter reference for the DB-2 parameters, making them get the values of the im-parameters when the button is clicked, like so:


                                produce_name - {im_product_name}

                                start_ww - {im_start_ww}

                                end_ww - {im_end_ww}


                                Best regards, Theo Klemming