A record in a table contains a range of valid dates, say:
*tbl1.start_date* and *tbl1.end_date*. So to ensure I get all records that are valid for a specific date range, the selection logic is: <...> WHERE end_date >= #dtFrom AND start_date < #dtTo (the #dtTo parameter used in the SQL statement is actually the calculated next day of the *#prmDt_To* parameter used in the report).
Now in a report I need to count the number of records for each day within the specified data range and include the days, if any, for which there were no valid records. Thus a retrieved record may be counted in several different days. I can do it relatively easily with a recursive CTE within the data set, but my rule of thumb is to avoid the unnecessary load on the SQL database and instead return just the necessary raw data and let the Report engine handle groupings. So is there a means to do this within SSRS?
Thank you,
Sergey
You might be able to do something in SSRS with custom code, but I recommend against it. The place to do this is in the dataset. SSRS is not designed to fill in groups that don't exist in the dataset. That sounds like what you are trying to do: SSRS would need to create the groups for each date whether or not that date is in the dataset.
If you don't have a number or date table in your database, I would just create a recursive CTE with a record for every date in the range that you are interested as you mention. Then outer join this to your table and use COUNT(tbl1.start_date) to find the appropriate days. This shouldn't be too painful a query for SQL server.
If you really need to avoid the CTE, then I would create a date or number table to use to generate the dates in your range.
Related
Require assistance in calculating the Total Active Users from March 16 2020 to Feb 16 2020.
I have tried using calculated fields, but not getting the correct results. Please advise.
Thank you,
Nirmal
To find the number of unique values that appear in a field, say [user_code], you can use the COUNT DISTINCT function, COUNTD() as in COUNTD([user_code])
To restrict the data to a particular time range, one way is put your date field on the Filter shelf and choose the settings that include only the data rows you want — say the range from 2/16 to 3/16 as you stated.
Alternatively, you can push the filtering condition into the calculation with an IF function call, as in COUNTD(IF <data is relevant> THEN [user_code] END) Thus effectively combining the two techniques. That works because if there is no ELSE clause and the IF condition is False then the IF statement evaluates to null. Since COUNTD() silently ignores nulls, like other aggregation functions, the expression acts as if the irrelevant data rows were filtered.
So, for example,
COUNTD(IF [dates] >= #2/16/2020# AND [dates] <= #3/16/2020# THEN [user_code] END)
Will tell you then number of unique user codes during the period between 2/16 and 3/16. The DateDiff() function will probably be useful in more elaborate tests.
Finally, what if you want more flexibility? You could easily use Parameters or Filter controls to let the user choose the date range interactively.
If you want this calculation repeated for each possible day, showing the unique users in the preceding 30 day period, as some sort of rolling calculation, then you’ll need to learn about some more advanced features. Either multiple calculations as above for different time ranges, using Table Calculations, or some data prep and/or data padding with Tableau Prep Builder, Python or some other technique — mostly because in that scenario each data row contributes to multiple rolling counts, rather than one count when partitioning the data by some dimension.
I have a data set that we are pulling from two different SQL tables. This is a custom SQL query. This is financial data - trades and transactions. I am pulling them together in a single return so for each customer I can see all the trades and transactions in a single query. To be specific - some trades do not have corresponding transactions, and some transactions do not have corresponding trades - therefore there are some rows that will have both a trade AND transaction date, some will have one or the other. For those missing they are NULL.
My question is - how can I create a filter so that I get all trades and transactions in a given date range? I tried using two filters (one for each date), but for some reason that won't work. Also, I'd like for our customers to be able to have one filter that gives them any trade or transaction in that range.
I thought about a parameter, but with the data always changing (this gets refreshed daily), I need something dynamic - one filter that will filter the data set on those two fields. Any suggestions would be greatly appreciated. Thanks so much!!
I have a table of users and another table of transactions.
The transactions all have a date against them. What I am trying to ascertain for each user is the average time between transactions.
User | Transaction Date
-----+-----------------
A | 2001-01-01
A | 2001-01-10
A | 2001-01-12
Consider the above transactions for user A. I am basically looking for the distance from one transaction to the next chronologically to determine the distances.
There are 9 days between transactions one and two; and there are 2 days between transactions three and four. The average of these is obviously 4.5, so I would want to identify the average time between user A's transactions to be 4.5 days.
Any idea of how to achieve this in Tableau?
I am trying to create a calculated field for each transaction to identify the date of the "next" transaction but I am struggling.
{ FIXED [user id] : MIN(IF [Transaction Date] > **this transaction date** THEN [Transaction Date]) }
I am not sure what to replace this transaction date with or whether this is the right approach at all.
Any advice would be greatly appreciated.
LODs dont have access to previous values directly, so you need to create a self join in your data connection. Follow below steps to achieve what you want.
Create a self join with your data with following criteria
Create an LOD calculation as below
{FIXED [User],[Transaction Date]:
MIN(DATEDIFF('day',[Transaction Date],[Transaction Date (Data1)]))
}
Build the View
PS: If you want to improve the performance, Custom SQL might be the way.
The only type of calculation that can take order sequence into account (e.g., when the value for a calculated field depends on the value of the immediately preceding row) is a table calc. You can't use an LOD calc for this kind of problem.
You'll need to understand how partitioning and addressing works with table calcs, along with specifying your sort order criteria. See the online help. You can then do something like, for example, define days_since_last_transaction as:
if first() > 0 then min([Transaction Date]) -
lookup(min([Transaction Date]), -1) end
If you have very large data or for other reasons want to do your calculations at the database instead of in Tableau by a table calc, then you use SQL windowing (aka analytical) queries instead via Tableau's custom SQL.
Please attach an example workbook and anything you tried along with the error you have.
This might not be useful if you cannot set User ID Field as a filter.
So, you can set
User ID
as a filter. Then following the steps mentioned in here will lead you to calculating difference between any two dates. Ideally if you select any one value in the filter, the calculated field from the link should give you the difference in the dates that you have in the transaction dates column.
I'm trying to annualise my data in tableau, but get an error in the Calculated Field.
"Cannot mix aggregate and non-aggregate arguments to function"
my formula is
sum(profit)/month(selected date) *12
How do I get an integer for the current month? That seems to be the problem, it tries to aggregate the month as well.
Thanks.
Short answer: wrap the call to month in a call to min() -- which works well if you have MONTH([selected date]) on the visualization as a dimension.
There are three types of calculated fields in Tableau:
row level calculations which act on a single data row. They can read from values of other fields in the same row and return a single value per row.
aggregate calculations which act on a partition or block of data rows. They can reference the result of aggregating the values for a field across the entire partition, using a an aggregate function like SUM() or MIN().
table calculations which act on an entire table of aggregated results.
You can't mix and match. Everything in a calculated field must be all at one level or another -- either all referenced fields must use aggregation functions (for aggregate calculated fields) or no referenced fields must use aggregation functions (for data row level calculated fields).
Hence the error message you saw.
Sometimes you know that all values for a field will be the same in a partition based on your visualization, so the aggregation function seems unnecessary. But Tableau still requires you to be explicit about how to turn a block of values into a single value, because the calculation must be defined even when the visualization is partitioned differently. In these cases, you can use min(), max(), avg(), or perhaps attr() because they all return the same value for a list of identical values.
The first two types are typically executed on the server (i.e. they are implemented by Tableau emitting SQL to send to the database server). Table calculations are executed by Tableau on the client site to post-process the results from the database server.
Table calcs are the most complicated type, but can be very useful. Explaining them is a post for another day.
I have used iReport to create a simple JasperReport which I run on a JasperServer. It queries some fields from a number of MySQL tables based on their creation timestamp. I am providing the start and end timestamps of the period to cover in the report as parameters of type java.sql.Timestamp. This works fine.
I was asked to introduce the ability to show a break-down on weeks or days of the report data. I would like to get some ideas on where to start with this. At this point I don't think I can accommodate this 'break-down' in the report query, since this feature seems beyond what SQL is designed to do. I know this sounds like an OLAP drill through, but I would like to avoid OLAP if possible (steep learning curve, tight deadlines).
My first thought was to create a subreport for each week or day . But this would leave me with an arbitrary number of subreports (depending on the overall time period covered by the report, which varies at each execution), and as far as I can tell iReport does not support this.
Here is one way to break the report down.
Create another parameter, groupby, which holds the a value that designates the grouping requirement. The values can be numeric, string or whatever else as long as it corresponds to day, week, month, etc.. grouping available.
Create a report group, breakdown, which will provide the breakdown. The group expression will depend on the groupby parameter. The expression is the date on the record except any detail finer than groupby value will be trimmed.
Create a variable, total, that will sum the data in the records. The variable should be reset on breakdown group and can be printed in the breakdown trailer band.
Make sure the sql queries sortby the date so that the groupby expression works.
Let me know if you have questions.