Selecting and Counting Specific Columns

[Eminence]

Hello,

Using SQLite, I was trying to select a specific column with having duplicate entries removed by using the DISTINCT function. 

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name) FROM data_db

What the above snippet does is just list only one supervisor but with the total count of unique employee names in the whole database when it should be listing out all supervisor entries from the database and list the total count of employees per each supervisors. If I remove the COUNT function, it does list out all unique entries of supervisor names from the list. Attached is a screenshot of an example database as well. 

Any help will be much appreciated. Thank you!

[Danp2]

A typical SQL Select statement would look like this --

SELECT  supervisor, COUNT(DISTINCT employee_name) FROM your_db GROUP BY supervisor

 

[Eminence]

3 hours ago, Danp2 said:

A typical SQL Select statement would look like this --

SELECT  supervisor, COUNT(DISTINCT employee_name) FROM your_db GROUP BY supervisor

 

There we go. Thanks a bunch! 

[Eminence]

Now I'm trying to count the number of employees per supervisor that has the attendance_status of Absent using this query.

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name), COUNT(attendance_status) FROM data_db WHERE attendance_status = 'ABSENT' GROUP BY supevisor"

But what it does is it also affects the second row(total count of the employees name) only showing employees that has the tagging of absent when I just want it to affect the third column. 

Edited August 8, 2018 by Eminence

[Danp2]

There are several different ways to do this. Here's a link that give some possible solutions.

https://stackoverflow.com/questions/19046812/multiple-count-for-multiple-conditions-in-one-query-mysql

[Eminence]

3 hours ago, Danp2 said:

There are several different ways to do this. Here's a link that give some possible solutions.

https://stackoverflow.com/questions/19046812/multiple-count-for-multiple-conditions-in-one-query-mysql

Thanks! Will check on this. 

[Eminence]

Just another quick question related to this, do I need to have a separated column in the database if I want to calculate the difference of the count of the total employees' schedule status tagged as work and the count of the total employees' attendance status tagged as absent?

What I want to do is calculate the absenteeism rate per supervisor based on the formula above however, I don' have a column for the absenteeism rate.

[Danp2]

If I understand correctly, then you should be able to perform a calculation as part of the overall query. Can you show us your existing SQL statement?

[Eminence]

Below is my current SQL statement. I'm still figuring out how to calculate the percentage of the total count of scheduled vs absent employees though.

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name), COUNT(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 END) AS count1, COUNT(CASE WHEN late_hours > 0.00 THEN 1 END) AS count1 FROM data_db WHERE supervisor != '' GROUP BY supervisor"

 

[Danp2]

You have two different fields named 'count1'. Not sure what's up with that. I would try something like this --

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name) as emp_count, COUNT(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 END) AS abs_count, COUNT(CASE WHEN late_hours > 0.00 THEN 1 END) AS late_count, abs_count / emp_count * 100 AS abs_rate FROM data_db WHERE supervisor != '' GROUP BY supervisor"

 

[Eminence]

2 hours ago, Danp2 said:

You have two different fields named 'count1'. Not sure what's up with that. I would try something like this --

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name) as emp_count, COUNT(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 END) AS abs_count, COUNT(CASE WHEN late_hours > 0.00 THEN 1 END) AS late_count, abs_count / emp_count * 100 AS abs_rate FROM data_db WHERE supervisor != '' GROUP BY supervisor"

 

 

Thanks @Danp2. After running, it says no such column: abs_count, do I also have to create a column for abs_count in SQLite instead? 

[Danp2]

Not sure how to advise you from here. I believe the syntax I posted is correct. However, it's possible that SQLite doesn't support it.

 

[Eminence]

3 hours ago, Danp2 said:

Not sure how to advise you from here. I believe the syntax I posted is correct. However, it's possible that SQLite doesn't support it.

 

Could be. I'll look into it. Thanks again @Danp2! 

[jchd]

There is very little limit on the complexity of SQL requests in SQLite.

What remains unclear is the actual DB schema, the semantics of every column and the meaning of the data stored there.
Post the schema and a sample data so that we can build a request satisfying your requirements. Be as clear and formal as possible.

 

[Eminence]

2 hours ago, jchd said:

There is very little limit on the complexity of SQL requests in SQLite.

What remains unclear is the actual DB schema, the semantics of every column and the meaning of the data stored there.
Post the schema and a sample data so that we can build a request satisfying your requirements. Be as clear and formal as possible.

 

Here's what I have

 

Edited August 9, 2018 by Eminence
Corrected database screenshot

[jchd]

I don't get it. This isn't the schema (list of all DDL statements). Please list all "CREATE TABLE ..." statements.

What's the column "attendance_status" you're basing your request on?

[Eminence]

CREATE TABLE data_db (
    id                      INTEGER PRIMARY KEY AUTOINCREMENT,
    employee_number         STRING,
    employee_name           STRING,
    supervisor              STRING,
    date                    DATETIME,
    schedule_status         STRING,
    attendance_status       STRING,
    operational_minutes     STRING,
    late_hours              STRING
);

My bad. Here's the DDL. The data comes from a .CSV file generated from an internal web application.

So I'm trying to count the total number of employees' schedule_status tagged as WORK and their attendance_status tagged as ABSENT and from there, calculate the Abs. Rate %. 

Edited August 9, 2018 by Eminence

[Eminence]

So I tried this:

Dividing count of attendance_status tagged with absent over the count of shedule_status tagged with work.

Local $sQuery = "SELECT DISTINCT supervisor, COUNT(DISTINCT employee_name) as emp_count, COUNT(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 END) AS abs_count, COUNT(CASE WHEN late_hours > 0.00 THEN 1 END) AS late_count, (COUNT(CASE WHEN attendance_status IS 'ABSENT' THEN 1 END) / COUNT(CASE WHEN schedule_status IS 'WORK' THEN 1 END) * 100) FROM data_db WHERE supervisor != '' GROUP BY supervisor"

It does return a value of 0 though, whereas changing the order to schedule_status over the count of attendance_status it does return a different result. Any ideas?

[jchd]

Ok I believe I get the picture, more or less.

First there are some problems with your table design.

1) SQLite doesn't have datatypes STRING nor DATETIME. Those columns are parsed as being unknown datatypes with affinity BLOB. Probably not what you intended.
2) "operational_minutes" should probably call "operational_hours"
3) Both "operational_minutes" and "late_hours" should be REAL, not STRING (nor TEXT).
4) You can't make further use of result columns aliased. That's why "abs_count" is unknown.
5) You probably don't need to use "autoincrement" for the id primary key. That doesn't mean what you think it means.

To fix this I defined a table like this:

CREATE TABLE "data_db" (
  "id" INTEGER PRIMARY KEY,
  "employee_number" TEXT,
  "employee_name" TEXT,
  "supervisor" TEXT,
  "date" TEXT,
  "schedule_status" TEXT,
  "attendance_status" TEXT,
  "operational_hours" REAL,
  "late_hours" REAL);

I had to type (you pasted a picture) some random data in:

id  employee_number employee_name supervisor    date       schedule_status attendance_status operational_hours late_hours
1   100123          John Doe      Bryan McGrath 2018-08-09 WORK            WORK                           7.47       1.55
2   100124          Jane Doe      Bryan McGrath 2018-08-09 WORK            ABSENT                            0          0
3   178223          Joe Grifith   Neil Grant    2018-08-09 WORK            WORK                              8          0
4   182123          Ryan Banks    Summer Harley 2018-08-09 WORK            ABSENT                            0          0
5   143708          Paul Tanner   Summer Harley 2018-08-08 WORK            WORK                           7.15       0.34
6   075891          Bill Moots    Neil Grant    2018-08-09 WORK            WORK                           8.12          0
7   000156          Shirley Vroot Neil Grant    2018-08-07 WORK            ABSENT                            0          0
8   000163          Shirley Wok   Neil Grant    2018-08-07 WORK            WORK                           7.52          0
9   542980          Dan Vroot     Neil Grant    2018-08-07 WORK            WORK                           6.45          0
10  514802          Sheila Vroot  Neil Grant    2018-08-07 WORK            WORK                           8.33       1.06

Then this statement should be close to what you want:
SELECT
  supervisor,
  COUNT(employee_name) emp_count,
  SUM(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 else 0 END) abs_count,
  SUM(CASE WHEN late_hours > 0 THEN 1 else 0 END) late_count,
  printf('%2.2f', 100.0 * SUM(CASE WHEN attendance_status LIKE 'ABSENT' THEN 1 else 0 END) / COUNT(employee_name)) abs_rate
FROM
  data_db
WHERE supervisor != ''
GROUP BY supervisor

The result then is:

supervisor      emp_count   abs_count   late_count  abs_rate
Bryan McGrath           2           1            1     50.00
Neil Grant              6           1            1     16.67
Summer Harley           2           1            1     50.00

Note that your DB currently contains text instead of numeric values.

[Eminence]

Thanks, @jchd! That's a great help, learned as well. It is working as intended now.