kdb+ - KDB fewer lines of code for modifying values in a col - Stack Overflow

admin管理员组

文章数量:1279207

I have a table like below:

q)tab:([]time:.z.t;sym:`ax`bx`cx`ds;col1:`a`a`b`b;price:1.4,1.2,1.3,1.4)
q)tab
time         sym col1 price
---------------------------
20:29:37.309 ax  a    1.4
20:29:37.309 bx  a    1.2
20:29:37.309 cx  b    1.3
20:29:37.309 ds  b    1.4
q)

My goal is to change values in col1 from a->b where a and b->a where b and have the same column name. Here is my code below that does the job but is there a better way to achieve this in fewer lines of code?

testTab:update col2:count[i]#enlist `b from tab where col1=`a
testTab:update col2:count[i]#enlist `a from testTab where col1=`b

q)testTab
time         sym col1 price col2
--------------------------------
20:29:37.309 ax  a    1.4   b
20:29:37.309 bx  a    1.2   b
20:29:37.309 cx  b    1.3   a
20:29:37.309 ds  b    1.4   a
q)testTab:`time`sym`col2 xcols testTab  //Bring forward col2

q)testTab
time         sym col2 col1 price
--------------------------------
20:29:37.309 ax  b    a    1.4
20:29:37.309 bx  b    a    1.2
20:29:37.309 cx  a    b    1.3
20:29:37.309 ds  a    b    1.4

q)testTab: delete col1 from testTab    //Delete original col1 from the table
q)testTab: `time`sym`col1 xcol testTab  // Rename col2 to col1

q)testTab
time         sym col1 price
---------------------------
20:29:37.309 ax  b    1.4
20:29:37.309 bx  b    1.2
20:29:37.309 cx  a    1.3
20:29:37.309 ds  a    1.4
q)

Thanks!

I have a table like below:

q)tab:([]time:.z.t;sym:`ax`bx`cx`ds;col1:`a`a`b`b;price:1.4,1.2,1.3,1.4)
q)tab
time         sym col1 price
---------------------------
20:29:37.309 ax  a    1.4
20:29:37.309 bx  a    1.2
20:29:37.309 cx  b    1.3
20:29:37.309 ds  b    1.4
q)

My goal is to change values in col1 from a->b where a and b->a where b and have the same column name. Here is my code below that does the job but is there a better way to achieve this in fewer lines of code?

testTab:update col2:count[i]#enlist `b from tab where col1=`a
testTab:update col2:count[i]#enlist `a from testTab where col1=`b

q)testTab
time         sym col1 price col2
--------------------------------
20:29:37.309 ax  a    1.4   b
20:29:37.309 bx  a    1.2   b
20:29:37.309 cx  b    1.3   a
20:29:37.309 ds  b    1.4   a
q)testTab:`time`sym`col2 xcols testTab  //Bring forward col2

q)testTab
time         sym col2 col1 price
--------------------------------
20:29:37.309 ax  b    a    1.4
20:29:37.309 bx  b    a    1.2
20:29:37.309 cx  a    b    1.3
20:29:37.309 ds  a    b    1.4

q)testTab: delete col1 from testTab    //Delete original col1 from the table
q)testTab: `time`sym`col1 xcol testTab  // Rename col2 to col1

q)testTab
time         sym col1 price
---------------------------
20:29:37.309 ax  b    1.4
20:29:37.309 bx  b    1.2
20:29:37.309 cx  a    1.3
20:29:37.309 ds  a    1.4
q)

Thanks!

• kdb+

Share Improve this question Follow asked Feb 24 at 20:37 CleanSockCleanSock 39522 gold badges44 silver badges1515 bronze badges

Add a comment  | 

1 Answer 1

Sorted by: Reset to default 3

Simplest way is a dictionary/map:

q)map:`a`b!`b`a
q)
q)update map col1 from tab
time         sym col1 price
---------------------------
21:01:27.484 ax  b    1.4
21:01:27.484 bx  b    1.2
21:01:27.484 cx  a    1.3
21:01:27.484 ds  a    1.4

If not all values are to change (i.e. only a subset are to be remapped) then only the subset that is changing should be in the map dictionary, and the rest can fill (^) with existing value

update col1^map col1 from tab

本文标签： kdbKDB fewer lines of code for modifying values in a colStack Overflow