Loading Profile...
Sharing a use-case
I thought I would share what I tried to do. Not sure how relevant the use-case is for your main target audience, anyway, maybe it could be considered for the future.
I wanted to upload historical stock prices for several stock symbols (and possibly some other data for the symbols - if multiple dimensions worked :)). The stock prices are available from Yahoo in form of CSV files, however it is one CSV file per stock symbol and the symbol itself is not part of the file (see http://ichart.finance.yahoo.com/table... as an example). I was hoping I could upload several files into the same dimension (doing append instead of replace) and also while uploading, specify another column manually - symbol name - with ability to set a value for each upload. Since this is not possible, I ended up doing some updates in a spreadsheet. Anyway, the data was too big and my spreadsheet could not handle merges of those large data, so after I added Symbol column to each of the CSV files, I had to use some other tool for merging. Bottom-line: "append" besides "update" (which replaces the data) would be nice. Also being able to add a column with manually specified value during the upload would be nice in this case - not sure if it would be useful in general though.
Also I noticed GoodData could be more generous about the formats it recognizes - e.g. date in a form of mm/dd/yy was not recognized as a date, column with numbers prefixed by $ was not recognized as countable (if not because of "$" then maybe because thousands were separated by comma).
Also, manually updating this sort of data is quite painful, so would be nice to have a way of automating that (see the other question I posted).
I wanted to upload historical stock prices for several stock symbols (and possibly some other data for the symbols - if multiple dimensions worked :)). The stock prices are available from Yahoo in form of CSV files, however it is one CSV file per stock symbol and the symbol itself is not part of the file (see http://ichart.finance.yahoo.com/table... as an example). I was hoping I could upload several files into the same dimension (doing append instead of replace) and also while uploading, specify another column manually - symbol name - with ability to set a value for each upload. Since this is not possible, I ended up doing some updates in a spreadsheet. Anyway, the data was too big and my spreadsheet could not handle merges of those large data, so after I added Symbol column to each of the CSV files, I had to use some other tool for merging. Bottom-line: "append" besides "update" (which replaces the data) would be nice. Also being able to add a column with manually specified value during the upload would be nice in this case - not sure if it would be useful in general though.
Also I noticed GoodData could be more generous about the formats it recognizes - e.g. date in a form of mm/dd/yy was not recognized as a date, column with numbers prefixed by $ was not recognized as countable (if not because of "$" then maybe because thousands were separated by comma).
Also, manually updating this sort of data is quite painful, so would be nice to have a way of automating that (see the other question I posted).
1
person likes this idea
I like this idea!
Tell me when this idea gets some attention.
The more people who like this idea, the more it gets noticed.
The more people who like this idea, the more it gets noticed.
The company has this in progress.
Create a customer community for your own organization
Plans starting at $19/month
-
Inappropriate?Yes, I agree the upload service still has some rough edges and you need to do some pre-processing in other tools. We certainly want to handle the use-cases that you are talking about.
First, we are working on the incremental data load just now. This means that we should be able to release it in January. Unfortunately we have postponed the data cleansing capability to the next sprint. The data cleansing introduces various pre-packaged transformations. For example removing extra characters ('$',',' etc.) from numbers or converting various date formats to the ISO date format that our backend supports.
We did not think about the "default" value for a certain column (your stock symbol). I like this transformation. I'll discuss this with our team here.
Thank you for very good feedback!
EMPLOYEE
