I have 3 columns (DateTime, GroupName, Value), some of these groups are closely related and I would like to display these in a stacked graph. The problem I am facing (I THINK) is that I don't have entries for all groups at all times.
(cannot find a decent way to add a table, so here is some code)
Datetime Groupname Value
1 a whatever
1 b whatever
1 c whatever
2 a whatever
2 b whatever
3 a whatever
3 b whatever
3 c whatever
4 b whatever
So in the example I don't have an entry for C at time 2. And I also don't have an entry for A and B at time 4.
Resulting in:
edit: added to onedrive link
With my limited SQL skills I am not sure how to fix this. How do I get the graph to connect the points from the DateTime points where we do have data, and ignore DateTime points where we do not have data?
11-07-2016 Edit
Ok, so here some pictures of the actual data
No rep - Onedrive it is
https://1drv.ms/f/s!AhKMFQBAmZ7GgYEMzdpTBvuXTi5gAQ
the graph looks different than my first example because I set the X-axis to scalar.
On 7/5/2016 and 7/6/2016 (month/day/year notation) the CH3 is low and 0. If I remove CH3 from results the graph looks ok.
#sqlandmore.com
This is the query. Very basic.
Data is coming from a database were the datetime is not in a proper datetime format so thats gets converted into the wimsview.
SELECT
WimsView.TagID
,WimsView.SampleDateTime
,WimsView.SampleValue
,WimsView.TagName
FROM
WimsView
WHERE
WimsView.SampleDateTime > N'07/4/2016 00:00:00'
AND ((WimsView.TagName LIKE N'%MBA%') OR (WimsView.TagName LIKE N'%MBB%') OR (WimsView.TagName LIKE N'%GF1_DPC%')OR (WimsView.TagName LIKE N'%KF1_DPC%')OR (WimsView.TagName LIKE N'%CH3_DPC%'))
AND WimsView.SampleValue IS NOT NULL
You can't ignore specific values on the graph.
you can either change your select statement not to include them
you can calculate an "average" (if possible) for the missing value in order to fill the missing "points" in your graph
or another calculation (i.e - same value as previous one on the graph)
whatever you decide - it should be handled on a query level, not on a drwaing level
Related
I'm not sure if my question is really stupid, but I found nothing on the internet...
Is it possible to insert a specific value in a cell of a matrix?
for example I have a dataset like below:
Month Prod Amount
2 X 34$
11 Y 12$
7 Z 150$
and a matrix like:
-------| Month |
Prduct |SUM(Amount)|
So the row group are products and column group are the months of a specific year.
If I want to add an extra column, with a specific value chosen dynamically from the amount (for xample 150$) so to have
-------| Month |columnName
Prduct |SUM(Amount)| 150
is that possible? also if the value is repeated through the column (it would be useful if I wanted the new column to have this specific value added for each value)
thanks a lot!! :D
You can insert a value directly in your matrix but it will be repeated for each record.
The best way is to add a new column with conditional values is to do this in your dataset query. Probably with a CASE statement if you are using SQL.
EDIT: If you can't adjust the query for whatever reason, you can add the new column and use SWITCH function inside your textbox to achieve the same.
New user, so apologies in advance for bad formatting.
Essentially what I'm trying to do is be able to populate the staff_hours column where it equals NULL with the one value that IS NOT NULL. As you can see from the screenshot, there will only be one person who staffs an open cl_hole_staffing_no and as a result will have a start_dt (with time) and end_dt (with time) along with staff_hours. 16 people were offered a shift, and the person in row 15 accepted it is what is going on here.
The ideal output would be the staff_hours column is populated with the amount of time of the one person who ended up taking the open job, so 24.00 in this example. How can I write a formula to do this? I was thinking something like an array function in Excel, but am not sure how to do that in SQL.
Your explanation is a bit confusing about what you are really trying to achieve. However I think that what you really want is just to populate the staff_hours column, which can be achieved with the following:
UPDATE
your_table_name
SET
staff_hours = 24
WHERE
staff_hours is NULL;
EDIT
I get it now. You want to operate with the two dates and extract the amount of hours between them. Since you are in sql-server you can actually define a Computed Column in which you can use the values from other columns to compute the value you want.
You will need to create your table again. (The example below contains only the necessary attributes for it to work)
CREATE TABLE your_table_name
( id INT IDENTITY (1,1) NOT NULL
, staff_start_dt DATETIME
, staff_end_dt DATETIME
, staff_hours AS DATEDIFF(hh, staff_start_dt , staff_end_dt)
);
Now every time you insert a record on the table with both staff_start_dt and staff_end_dt, the column staff_hours will automatically compute the number of hours between the two dates.
[pre]
Code (vb):
A B C
1 10 X X
2 11 A Y
3 12 Y Z
4 13 B
5 14 B
6 15 Z
[/pre]
Assuming that the rows in Col A is Named "datarange"
And your criteria is in C1:C3
The following formula will return an array {10,12,15}
=SMALL(COUNTIF(C1:C3,B1:B6)*datarange, ROW(INDEX(A:A,SUMPRODUCT(--(COUNTIF(C1:C3,B1:B6)=0))+1):INDEX(A:A,ROWS(datarange))))
COUNTIF(C1:C3,B1:B6)*datarange returns {10;0;12;0;0;15}
The segment ROW(INDEX(....):INDEX(...)) returns {4;5;6}, indicating the number of non-zero values.
The SMALL() function then returns the 4th smallest, 5th smallest and 6th smallest values.
One disadvantage with this approach is that you get a sorted sub-list. Perhaps that would work for you.
I have this spreadsheet i am working on for awhile. Its basically attendance piece. End user keeps track of employees, if they showed up or not etc...
I have tired looking up loops but i couldn't figure out how to do what i am trying to do.
What i have in this excel.
A-D : Emp info
E-∞ : 1-3 Days/Dates; 4-∞ emp data (if they missed a day, values for that)
To get better understanding, see this
The data entered from E5 to xx thats where i am trying to get this vba working.
Anything the script detects first value either '1' or '2', start 90 days (cells) from there. And after 90, reset to 0. starting from 91 start search for '1' or '2' and do similar.
See the excel file for better understanding. If it doesn't make sense, ill be happy to simplify.
Thank You
The most efficient and clean way to handle this is to use a form of a relational data model because it can be done easily without using VBA code. You will have two simple tables in your spreadsheet, EmployeeInfo and AttendanceRecords. Your Employee info will look something like this
Emp# Name Craft # In 90 Days NumOf2s NumOf1s
1 EMP 1 SM Site Manager 0 0 0
2 EMP 2 SM Site Manager 1 0 1
3 EMP 3 SM Site Manager 0 0 0
4 EMP 4 SM Site Manager 0 0 0
5 EMP 5 SM Site Manager 1 0 1
The last three columns are calculated from the AttendanceRecords table. This table is going to be variable size but this way you only need to store the important data (When employees actually got marks). It will look like this.
Emp# Date Days Count
1 12/1/2013 122 1
3 1/1/2014 91 2
2 2/1/2014 60 1
5 2/15/2014 46 1
You can have multiple entries for the same day and the same employee. The important thing is that we only need one entry per infraction (NOTE: In order to do this in a proper database type model, each attendance record should also have some kind of incrementing totally unique ID (like employees), but we can forgo that for this application).
You enter in the employee number, the date, and the count. The "Days" column then auto calculates the age of the record with the following formula:
=TODAY()-[#Date]
NOTE: If the [#Date] notation does not look familiar, this is because it deals with Excel Tables. I recommend you read up on those if not already familiar.
So now we have the age of each record. So back on the EmployeeInfo table, we use the following formula to get all AttendanceRecords that apply to Employee x for the last 90 days
=SUMIFS(AttendanceRecords[Count],AttendanceRecords[Emp'#],[#[Emp'#]],AttendanceRecords[Days],"<=90")
You can now also use some simple formulas to get the other columns I pointed out, including the number of 2 count in fractions or the number of 1 count infractions:
=COUNTIFS(AttendanceRecords[Emp'#],[#[Emp'#]],AttendanceRecords[Days],"<=90",AttendanceRecords[Count],2)
=COUNTIFS(AttendanceRecords[Emp'#],[#[Emp'#]],AttendanceRecords[Days],"<=90",AttendanceRecords[Count],1)
There is a lot more data that could be gathered, including the date of the last infraction, total number of infractions for all time, etc. If any of the formulas or terms I used don't make sense or need more explaining, feel free to ask.
EDIT: If you want them automatically removed after 90 days, it would be relatively easy to write a VBA script to do this. It would also be easy to just sort the AttendanceRecords table on Days and delete all records that are older than a certain number of days. However, unless you see yourself adding hundreds of records a week, this really shouldn't be necessary. Also, If you want to write a Visual Basic form to enter in new infractions, that is definitely very possible, but another discussion.
EDIT: To respond to concerns about viewing when these issues happened, I will give you an example of a way to view the data in your tables. One of the advantages of excel tables is that the order of the records isn't as absolute as in a normal range, so we can sort, rearrange, and filter them to see what we need. So if you need to see all of the issues with employee 3, you just go to the Emp# column in the AttendanceRecords table, select the little arrow down button next to Emp#, uncheck 'Select All', and then check the '3', and then the only values I will see in the table are the ones from employee 3. I can then sort the 'Date' column by clicking its little arrow and selecting 'Sort Newest to Oldest'.
What it comes down to is that you can view ANY data you need to, and if you think through what you really need to see, you can set up your summary table (EmployeeInfo) to display enough data that you hardly ever have to look at the AttendanceRecords table. But if you need to, you can go into that table and do a manual sort (as I described above) very easily.
EDIT: To help add some of the functionality I've shown above to the askers current spreadsheet, I will show the current formula.
=SUMIFS($E5:$BR5,$E$3:$BR$3,">"&(TODAY()-90))
For EMP 1, this formula uses the employees row as the sum range. It then looks at the field of dates in the corresponding columns in row 3. If the date in row 3 is > TODAY()-90, then we will add it to the count. This will at least just look at the infractions for the previous 90 days.
I am creating a report using MS Report Builder 3.0. For this report, I have a stored procedure built that filters down to the specific rows needed, and then I use a row group to group on a particular field (pass_no). The table that is displayed is 2 columns and 3 rows within the row group. The basic description of what I want to accomplish is instead of the rows running onto the next page, I want the rows to continue on the same page in a new set of 2 columns. Think of it like a newspaper where the text continues in a new column rather than running down onto the next page.
For the example I'm going to use here, there are 12 rows of data returned by the SP, and 8 unique values in the pass_no column which is what my row group is grouped on. So in the report I end up with 8 groups of 3 rows. I'm aiming to have the table display 6 pass_no values (so 6 groups of 3 rows) before, for lack of better terminology, starting a new table.
My first approach at this has been to create a column group and set the grouping expression to the following:
=Floor((RowNumber(Nothing) - 1) / 6)
While this works in creating a new set of 2 columns, the split for the new columns is based on the row number from the raw data returned by the SP rather than the number of rows sets created by the row group. So because there are 12 rows returned, and the 6th and 7th rows have the same pass_no value, the second set of columns duplicates that 1 set of data. Also, the top 6 rows of the second column set are blank with the second set of values appearing below the first set.
If I add an additional column group where it is also grouping by pass_no, then I don't get the duplicate values, but I do get a pair of columns for each pass_no as well (as would be expected). I've tried modifying the expression above a bit and changed Nothing to the row group name and have tried the table name, but neither of them have yielded the desired result.
I can't alter the SP to do the grouping there because there are other column values that are not identical and I pull that data into a cell value expression within the table using Join(LookupSet()).
I have also considered creating 2 tables and applying a filter to the table so the first table only displays the first 6 results and the second table displays the remaining results, but that also looks at the raw data rather than the groupings and TOP N can't be used on pass_no as it's a text value, not an integer. This would also cause problems if I need to go to 3 tables.
So long story short, is there a way to do a table break rather than a page break or to overflow columns onto the same page rather than onto a new page?
Here's the pertinent portions of the Dataset:
http://sqlfiddle.com/#!2/5082b/1
PASS_NO MASTERTRAN TRANS_NO DESCRIPTION IS_MOD
7913019000 4931019000 4931019000 General Admission Adult 0
7914019000 4932019000 4932019000 Sea Turtle Hosp Adult 0
7914019000 4932019000 4933019000 2:00 PM SEA TURTLE HOSP 1
7916019000 4934019000 4934019000 Sea Turtle Hosp Child 0
7916019000 4934019000 4935019000 2:00 PM SEA TURTLE HOSP 1
7917019000 4934019000 4934019000 Sea Turtle Hosp Child 0
7917019000 4934019000 4935019000 2:00 PM SEA TURTLE HOSP 1
7918019000 4934019000 4934019000 Sea Turtle Hosp Child 0
7918019000 4934019000 4935019000 2:00 PM SEA TURTLE HOSP 1
7922019000 4936019000 4936019000 General Admission Child 0
7923019000 4936019000 4936019000 General Admission Child 0
7924019000 4936019000 4936019000 General Admission Child 0
I think your data presents a bit of a problem.
As you've already figured out, typically for this sort of setup you'd set up a row group with an expression like:
=(RowNumber(Nothing) - 1) Mod 6
And a column group expression like:
=Ceiling(RowNumber(Nothing) / 6)
This would create a six row tablix that would grow horizontally as required.
See this SO question for a similar example.
However, you currently have the requirement of also grouping by another column - pass_no in your case. Normally you can approximate a group-level row number with an expression like:
=RunningValue(Fields!pass_no.Value, CountDistinct, "DataSet1")
Unfortunately, when you try to add this into one of the grouping expressions like:
=Ceiling(RunningValue(Fields!pass_no.Value, CountDistinct, "DataSet1") / 6)
You get the following error:
A group expression for the tablix 'Tablix1' includes the aggregate
function RunningValue. RunningValue cannot be used in group
expressions.
Based on all this, my recommendation is to try and get a Dataset that has one row per pass_no value and base the tablix on this, with the above row/column grouping expressions, i.e. no need to group on multiple pass_no rows. So in your example it would have eight rows. You could then have a separate Dataset with all the individual rows and use a lookupset function to concatenate the description, etc.
Your other option is to try and get everything on one Dataset only, including the aggregates as required. This might not be possible, but for description at least you can leverage any of the various techniques here to get a delimited list. Once you have this list you can replace the delimiter with vbCrLf to split it back over multiple rows.
All this is a very long-winded way of saying that I don't know if your requirement is possible with your data, but if you look at having at least one Dataset with one row per pass_no you should be able to make it work.
Anyone have advice on how to build an average measure that is dynamic -- it doesn't specify a particular slice but instead uses your current view? I'm working within a front-end OLAP viewer (Strategy Companion) and I need a "dynamic" implementation based on the dimensions that are currently filtered in the data view.
My fact table looks something like this:
Key AmountA IndicatorA AmountB Other Data
1 5 1 null 25
2 6 1 null 52
3 7 1 2 106
4 null 0 4 108
Now I can specify a simple average for "[Measures].[AmountA]" with "[Measures].[AmountA] / [Measures].[IndicatorA]" which works great - "[IndicatorA]" sums up to the number of non-null values of "[AmountA]". And this also works great no matter what dimensions are selected in the view - it always divides by the count of rows that have been filtered in.
But what about [AmountB]? I don't have a null indicator column. I want to get an average value of [AmountB] for whatever rows have been filtered in for my current view. If I try to use the count of rows as a simple formula (psuedo-code "[Measures].[AmountB] / Count([Measures].[Key])") I get the wrong result, because it is counting all the null rows in the average.
So, I need a way to use the AVG function to specify the average of [AmountB] over the set of "whatever rows I'm currently filtering in, based on whatever dimensions I'm currently using". How do I specify this dynamic set?
I've tried several different uses of the AVG function and they have either returned null or summed up to huge numbers, clearly not the average I'm looking for.
Thanks-
Matt
Sorry, my first suggestion was wrong. If you don't have access to OLAP cube you can't write any mdx-query for this purpose (IMHO). Because, you don't have any detailed data (from your fact table) in this access level and you can use only aggregated data and dimensions from your cube.
Otherwise (if you have access to olap db), you can create this metric (count of not NULL rows) in your measure group and after that use it for AVG calculation (as calculated member in your cube or in section "WITH" in your mdx-query).