Get DATE when Price Changed - sql

Question

Im trying to perform some DATEDIFF calculations but i simply lack the knowledge to complete the task.
I got a table that keeps record of the articles sold with date when the operation took place and historic prices of the article. Since i live in a country with 30% of anual inflation. I must consider that factor to calculate the earning by every article sold.
It all works fine while i have the buying price saved and the date when a product price change took place. But, when that data was not saved as an approximation i would like to use the date when the operation with a different price took place.
As result of the following select operations:
USE Rusich
DECLARE
#IDNegocio AS INT,
#UsarIDProducto AS BIT,
#IDProducto AS VARCHAR(50),
#FechaInicio AS DATE,
#FechaFinal AS DATE,
#TamañoMinimoDeMuestra AS INT,
#SeVendeMinimoCadaXDias AS INT,
#Inflacion AS DECIMAL(18,3),
#AuxMargen AS DECIMAL(18,2),
#AuxDifDias AS INT;
SET #FechaInicio = '01/02/10';
SET #FechaFinal = '28/02/18';
SET #IDNegocio = 1;
SET #IDProducto = '6910101404918'; /*<<<< Cambiar a NULL segun necesidad*/
SET #TamañoMinimoDeMuestra = 1;
SET #SeVendeMinimoCadaXDias = 90;
SET #Inflacion = 0.083;
SELECT
AUXFechas.IDProducto,
AUXFechas.[$ Precio de Venta Historico],
AUXFechas.[Fecha Precio],
AUXFechas.[Fecha Venta],
AUXFechas.FechaAnterior,
AUXFechas.DiasDiferencia,
CAST(
CASE WHEN
AUXFechas.IDProducto NOT IN ('112', '113' , '114')
AND AUXFechas.[Fecha Precio] <= AUXFechas.[Fecha Venta]
AND StockDetalles.[Precio de Compra] <> 0
AND StockDetalles.[Precio de Compra] IS NOT NULL THEN
(1 - StockDetalles.[Precio de Compra] / AUXFechas.[$ Precio de Venta Historico]) * 100 - (AUXFechas.[$ Descuento]/100) - DATEDIFF(DAY, AUXFechas.[Fecha Precio], AUXFechas.[Fecha Venta]) * #Inflacion
ELSE
NULL
END
AS DECIMAL(18,2)) AS [% Margen Ganancia Real]
FROM
StockDetalles JOIN (
SELECT
Stock.ID AS IDProducto,
Stock.Cantidad AS #Disponible,
Recibos.Cantidad as [#Vendida],
Recibos.Precio AS [$ Precio de Venta Historico],
Recibos.Descuento AS [$ Descuento],
Stock.[Precio de Venta] AS [$ Precio],
Stock.[Fecha Actualizacion de Precio] AS [Fecha Precio],
CAST(RecibosRegistros.Fecha AS DATE) AS [Fecha Venta],
ISNULL(LAG(CAST(RecibosRegistros.Fecha AS DATE), 1) OVER (PARTITION BY Recibos.IDProducto ORDER BY RecibosRegistros.Fecha), Stock.[Fecha Actualizacion de Precio]) AS FechaAnterior,
/*Considera el caso cuando no existe un registro anterior y usa la fecha de actualizacion de precio si es posible*/
CASE WHEN (LAG(CAST(RecibosRegistros.Fecha AS DATE), 1) OVER (PARTITION BY Recibos.IDProducto ORDER BY RecibosRegistros.Fecha) IS NULL) AND CAST(RecibosRegistros.Fecha AS DATE) > Stock.[Fecha Actualizacion de Precio] THEN
(DATEDIFF(DAY, CAST(RecibosRegistros.Fecha AS DATE),
/*PARTITION BY PARTICIONA POR COLUMNA*/
ISNULL(LAG(CAST(RecibosRegistros.Fecha AS DATE), 1) OVER (PARTITION BY Recibos.IDProducto ORDER BY RecibosRegistros.Fecha), [Fecha Actualizacion de Precio])) / Recibos.Cantidad) * -1
ELSE
(DATEDIFF(DAY, CAST(RecibosRegistros.Fecha AS DATE),
/*PARTITION BY PARTICIONA POR COLUMNA*/
LAG(CAST(RecibosRegistros.Fecha AS DATE), 1) OVER (PARTITION BY Recibos.IDProducto ORDER BY RecibosRegistros.Fecha)) / Recibos.Cantidad) * -1
END AS DiasDiferencia
FROM
RecibosRegistros
JOIN Recibos ON RecibosRegistros.IDRecibo = Recibos.IDRecibo
JOIN Stock ON Recibos.IDProducto = Stock.ID
WHERE
RecibosRegistros.NegocioID = #IDNegocio
AND Stock.IDNegocio = #IDNegocio
AND Stock.ID != '111'
AND Stock.ID != '112'
AND Stock.ID != '113'
AND Stock.ID != '114'
AND Stock.ID = COALESCE(#IDProducto,Stock.ID)) AS AUXFechas ON StockDetalles.ID = AUXFechas.IDProducto
JOIN Categorias ON StockDetalles.CategoriaID = Categorias.ID
Im obtaining this result:
IDProducto $ Precio de Venta Historico Fecha Precio Fecha Venta FechaAnterior DiasDiferencia % Margen Ganancia Real
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
6910101404918 15,00 2012-10-18 2012-07-23 2012-10-18 NULL NULL
6910101404918 15,00 2012-10-18 2012-09-01 2012-07-23 40 NULL
6910101404918 21,00 2012-10-18 2013-07-01 2012-09-01 303 NULL
6910101404918 21,00 2012-10-18 2013-09-02 2013-07-01 63 NULL
6910101404918 21,00 2012-10-18 2013-09-24 2013-09-02 22 NULL
0200002003867 28,00 2014-03-05 2014-07-08 2014-03-05 125 NULL
0200002003867 28,00 2014-03-05 2014-07-08 2014-07-08 0 NULL
0200002003867 28,00 2014-03-05 2014-10-23 2014-07-08 107 NULL
0200002003867 28,00 2014-03-05 2015-01-21 2014-10-23 90 NULL
0200002003867 28,00 2014-03-05 2015-04-06 2015-01-21 75 NULL
What i need to do is get a column with date (Fecha Precio) where [$ Precio de Venta Historico] changed.
Output should be:
IDProducto $ Precio de Venta Historico Fecha Precio Fecha Venta FechaAnterior DiasDiferencia % Margen Ganancia Real Fecha Cambio
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
6910101404918 15,00 2012-10-18 2012-07-23 2012-10-18 NULL NULL NULL
6910101404918 15,00 2012-10-18 2012-09-01 2012-07-23 40 NULL 2012-10-18
6910101404918 21,00 2012-10-18 2013-07-01 2012-09-01 303 NULL NULL
6910101404918 21,00 2012-10-18 2013-09-02 2013-07-01 63 NULL 2012-09-01
6910101404918 21,00 2012-10-18 2013-09-24 2013-09-02 22 NULL 2012-09-01
0200002003867 28,00 2014-03-05 2014-07-08 2014-03-05 125 NULL NULL
0200002003867 28,00 2014-03-05 2014-07-08 2014-07-08 0 NULL 2014-07-08
0200002003867 28,00 2014-03-05 2014-10-23 2014-07-08 107 NULL 2014-07-08
0200002003867 28,00 2014-03-05 2015-01-21 2014-10-23 90 NULL 2014-07-08
0200002003867 28,00 2014-03-05 2015-04-06 2015-01-21 75 NULL 2014-07-08
My guess is that the solution must be related with LAG/LEAD and PartitionBY IDProducto but i dont know how to do it. Please Help

Answer 1

Added this code to try to get some result and i almost got it but not quite yet:
LAG(AUXFechas.[Fecha Venta], 1, NULL) OVER (PARTITION BY IDProducto, AUXFechas.[$ Precio de Venta Historico] ORDER BY AUXFechas.[Fecha Venta])
GOT:
6910101404918 Bandeja Plastica Borde Dorado Oval 23 Bazar 0 1 15,00 21,00 2012-10-18 2012-07-23 2012-10-18 NULL NULL NULL
6910101404918 Bandeja Plastica Borde Dorado Oval 23 Bazar 0 1 15,00 21,00 2012-10-18 2012-09-01 2012-07-23 40 NULL 2012-07-23
6910101404918 Bandeja Plastica Borde Dorado Oval 23 Bazar 0 1 21,00 21,00 2012-10-18 2013-07-01 2012-09-01 303 NULL NULL
6910101404918 Bandeja Plastica Borde Dorado Oval 23 Bazar 0 1 21,00 21,00 2012-10-18 2013-09-02 2013-07-01 63 NULL 2013-07-01
6910101404918 Bandeja Plastica Borde Dorado Oval 23 Bazar 0 1 21,00 21,00 2012-10-18 2013-09-24 2013-09-02 22 NULL 2013-09-02
That last date (2013-09-02) Should be the same as previous (2013-07-01)
Tryed with another product and got:
0200002003867 Bandeja Acero Oval 21cm Bazar 0 1 28,00 28,00 2014-03-05 2014-07-08 2014-03-05 125 NULL NULL
0200002003867 Bandeja Acero Oval 21cm Bazar 0 1 28,00 28,00 2014-03-05 2014-07-08 2014-07-08 0 NULL 2014-07-08
0200002003867 Bandeja Acero Oval 21cm Bazar 0 1 28,00 28,00 2014-03-05 2014-10-23 2014-07-08 107 NULL 2014-07-08
0200002003867 Bandeja Acero Oval 21cm Bazar 0 1 28,00 28,00 2014-03-05 2015-01-21 2014-10-23 90 NULL 2014-10-23
0200002003867 Bandeja Acero Oval 21cm Bazar 0 1 28,00 28,00 2014-03-05 2015-04-06 2015-01-21 75 NULL 2015-01-21
Only the first cell after NULL in the right column is OK but the rest sould be the same

Answer 2

Finally got what i wanted by using the FIRST_VALUE function:
FIRST_VALUE(AUXFechas.[Fecha Venta]) OVER (PARTITION BY AUXFechas.IDProducto, AUXFechas. [$ Precio de Venta Historico] ORDER BY AUXFechas.[Fecha Venta])
AND GOT THIS!:
IDProducto Fecha Precio Fecha Venta FechaAnterior DiasDiferencia (Sin nombre de columna)
7790002000483 2014-08-28 2012-04-14 2014-08-28 NULL 2012-04-14
7790002000483 2014-08-28 2012-05-03 2012-04-14 19 2012-04-14
7790002000483 2014-08-28 2012-05-03 2012-05-03 0 2012-04-14
7790002000483 2014-08-28 2012-06-02 2012-05-03 30 2012-04-14
7790002000483 2014-08-28 2012-06-02 2012-06-02 0 2012-04-14
7790002000483 2014-08-28 2012-06-02 2012-06-02 0 2012-04-14
7790002000483 2014-08-28 2012-06-18 2012-06-02 16 2012-04-14
7790002000483 2014-08-28 2012-07-05 2012-06-18 17 2012-04-14
7790002000483 2014-08-28 2012-07-14 2012-07-05 4 2012-04-14
7790002000483 2014-08-28 2012-07-20 2012-07-14 6 2012-04-14
7790002000483 2014-08-28 2012-07-23 2012-07-20 3 2012-04-14
7790002000483 2014-08-28 2012-07-23 2012-07-23 0 2012-04-14
7790002000483 2014-08-28 2013-01-30 2012-07-23 191 2013-01-30
7790002000483 2014-08-28 2013-07-20 2013-01-30 171 2013-01-30
7790002000483 2014-08-28 2013-10-16 2013-07-20 88 2013-01-30
7790002000483 2014-08-28 2015-07-14 2013-10-16 636 2015-07-14