For two tables each with datetime and ticker symbol columns, how can we achieve the functionality of the binary function bin within each ticker group. That is, instead of returning the latest index from the entire left-table prior to the time of each right-table row, for a given right-table row it should return the latest index from the left-table amongst only the rows of the same ticker symbol as the right-table row.
My first thought would be to add a per-group index in the left-table, apply bin on each ticker group for it’s group-index and then use the unique (ticker,group-index) pair to find the index on the full left-table. However, I am not sure how to implement this or if this is the best way to achieve the desired functionality.
Could you give some sample inputs and desired output?
This sounds like something you can solve with aj
Check https://code.kx.com/q/ref/aj/ for details
Related
I have a task to create tetrapeptide screening library aminoacids using Knime. I have never used Knime before sadly. I need to create a workflow with all 20 aminoacids, multiply it with another 20, then multiply the result with another 20 and repeat to get final result of tetrapeptides. Can someone suggest me how to input aminoacids on the Knime? Thank you very much!
Use a Table Creator node to enter the Amino acid single-letter codes, one per table. Now use a Cross Joiner node to cross-join the table to itself - you should now have a table with rows like:
A|A
A|C
etc.
Now put this table into both inputs of a second Cross Joiner node, which should give you now quite a long table starting something like:
A|A|A|A
A|A|A|C
A|C|A|A
A|C|A|C
etc.
Now use a Column Aggregator node, select all column as aggregation columns, the aggregation method as Concatenate and change the delimiter to an empty string:
and:
This will give you a table with a single column, 'Peptide':
AAAA
AAAC
ACAA
ACAC
etc.
If you want the output as a chemical structure, then as of v1.36.0 the Vernalis community contribution contains a node Speedy Sequence to SMILES which will convert the sequence to a SMILES string (make sure you select the option that your input column is a Protein!)
The full workflow is as shown:
I am trying to add a column to a collection by multiplying the 0.9 to existing database column recycling. but I get a run time error.
I tried to multiply 0.9 direction in the function but it is showing error, so I created the class and multiplied it there yet no use. what could be the problem?
Your error message is telling you what the problem is: your database query is using GROUP BY in an invalid way.
It doesn't make sense to group by one column and then select other columns (you've selected all columns in your case); what values would they contain, since you haven't grouped by them as well (and get one row returned per group)? You either have to group by all the columns you're selecting for, and/or use aggregates such as SUM for the non-grouped columns.
Perhaps you meant to ORDER BY that column (orderBy(dt.recycling.asc()) if ascending order in QueryDSL format), or to select all rows with a particular value of that column (where(dt.recycling.eq(55)) for example)?
I have a problem with Tableau.
I have a Dataset with some rows and some columns. I want to write a IFELSE structure where
The IF condition is that the Value of special field(fixed by a row and a column) is equal to the header of a column (it is every time equal to one name(header) of the different columns).
So to summarize: one value is every time equal to the name of a column and to find the column shall be the if-structure
Does someone know if there is a function to call the name (header) of a column? I didn't find it
Here is an small example, in which the Calculated_function choose the right price according to the Barcode. Everything in the first raw, is the header_name of the column below. enter image description here
Best regards
Jonas
You can work like this.
I created a sample dataset as given by you
Step-1: Connected With data in tableau. Clicked all columns having price (4 here), pivoted them so that they look like this..
Step_2: Create calculated_field like this
if [Barcode] = [Barcode_c]
then [Price] END
Step3: Filtered out null values from calculatedField and got a view like this which can be tweaked as per liking.
I am trying to filter a table visualization of all of my data by looking to see if a Study Number contains Activity A. If a Study Number contains Activity A then I want to filter for all rows containing those Study Numbers even if the Activity is not A. See mock data below. In my real data set I have ~55,000 rows.
I have created a calculated to return Study Numbers if Activity= A but I am not sure where to go from there. Thanks for any help.
If(UniqueConcatenate([Activity]) OVER ([Study Number])~="A","Y","N")
Will give you this resulting column that you can then filter on (or you can use the formula as a Data Limiting Expression:
I am working on data in Spotfire. The table has 4 columns:
RowID
StudID
IMT
Date
I am trying to insert a calculated column in Spotfire to get the date from the previous row for a specific StudID. The date should not be filled for first entry for a specific StudID since it does not have a previous row.
Please refer to the image for details:
This will be a calculated column using the OVER function, along with Intersect, Previous and the First aggregation.
First([Date]) OVER Intersect(Previous([Date]), [StudID])
It reads: over the intersection between (group of) the previous (to the current row) dates (which are the same) and the Student ID's (the same as the current row), give me the first row of that group. In your example, it will only ever return one date for that group, but the formula needs to be able to handle what happens if there are multiple rows. You may also need to think about whether this will happen in your data and what you're going to do about it. I.e.
StudID Date
124-639 6/12/2018
124-639 6/12/2018
124-639 6/14/2018
Building off of JasonJ's answer, it looks like his solution ran into issues when the dates of different StudIDs overlapped with one another.
So I was seeing something along the lines of this:
StudID, Date, Result
A, 10/1/2014,
A, 10/10/2014, 10/1/2014
A, 10/17/2014, 10/10/2014
B, 10/20/2014,
A, 10/21/2014,
B, 10/22/2014,
B, 10/24/2014, 10/22/2014
I created a weird workaround by adding another Calculated Column.
I doubt this is the IDEAL way to do this (I'd bet there's a better OVER function, but I couldn't identify it right off), but it looks like it's working.
First Calculated Column (Named [CalcRank]):
Rank(Concatenate([StudID],Year([Date]),If(DayOfYear([Date])<10,"0",""),If(DayOfYear([Date])<100,"0",""),DayOfYear([Date])))
Second Calculated Column:
Max([Date]) OVER (Intersect(Previous([CalcRank]),[StudID]))
Please note, you may have to pad your StudID with 0s to make sure it orders properly, like I did with the Date column.