Data needs more cleaning

[snowch]

Looking at the description, it appears that they have discarded the stock. I think it doesn't have a 'C' because these were never part of a customer order. This is my best guess though :)

InvoiceNo	StockCode	Description	Quantity	InvoiceDate	UnitPrice	CustomerID	Country
556690	23005	printing smudges/thrown away	-9600	14/06/2011 10:37	0		United Kingdom
556691	23005	printing smudges/thrown away	-9600	14/06/2011 10:37	0		United Kingdom

Now that I've had a better look through the data, I think I should ignore rows with negative values where the Invoice number doesn't start with 'C'. I think I should probably ignore all records that have a UnitPrice of 0.

Also, where there is a 'C' record, I think I should ignore the corresponding record for the invoice like here:

InvoiceNo	StockCode	Description	Quantity	InvoiceDate	UnitPrice	CustomerID	Country
C581484	23843	PAPER CRAFT , LITTLE BIRDIE	-80995	09/12/2011 09:27	2.08	16446	United Kingdom
581483	23843	PAPER CRAFT , LITTLE BIRDIE	80995	09/12/2011 09:15	2.08	16446	United Kingdom

I have a feeling that the above order may have been cancelled because the quantity entered was a mistake. It looks like the same may have been the case for:

InvoiceNo	StockCode	Description	Quantity	InvoiceDate	UnitPrice	CustomerID	Country
C541433	23166	MEDIUM CERAMIC TOP STORAGE JAR	-74215	18/01/2011 10:17	1.04	12346	United Kingdom
541431	23166	MEDIUM CERAMIC TOP STORAGE JAR	74215	18/01/2011 10:01	1.04	12346	United Kingdom