ZD's dashboard

What browser do you use? Can you take a look at this post http://getsatisfaction.com/gooddata/t... ?

First, thank you for the excellent feedback!

The COUNT metrics are a bit more tricky. We are still scratching our heads how to make them easier to use.

Let me start with a simple visualization of your model. The picture below shows a structure of your data.



The COUNT aggregation function can accept one or two arguments. The first one is what you want to count. You are counting distinct values of an attribute (non-countable field). You choice was right, you used the ID. As the ID is unique for each call, the COUNT(ID) computes the total count of all calls. This number is always the same, no matter in what report you place it. No matter how you filter the calls (e.g. by PHONE # or EMPLOYEE). The COUNT focuses solely on the ID and always returns the same number.

Then you wanted to do something a bit different. Let's say that want to count calls from specific employee. Then you need to explain what you want to count if you put an EMPLOYEE.NAME together with COUNT(ID) to a report. In your case, you want the system to count the number of calls that the employee made. This is the point when you need to use the second COUNT function argument. Our system creates a special "Facts of X" attribute that represents an entity (e.g. calls, opportunities etc.). This is exactly the attribute that you need to use to tell the system, that you want it to count calls when you put COUNT(ID) and EMPLOYEE.NAME in your report. So the final formula is COUNT(ID,Facts of CALL).

You need to use the Advanced metric editor to create such metric. Please click advanced link in the WHAT part of the Slice & Dice and then follow the Custom metric link, give your metric a name (e.g. Count of Calls) and write the metric definition: SELECT COUNT(ID,Facts of CALL) . Please note that you'll need to select the attributes ID and Facts of CALL from the right list (you still can't type the names unfortunately). I would recommend to make this metric global in order you can reuse it in your other reports.

Regarding the mapping of multiple dates in our UI. This is something what we are working on right now. You should be able to do such multiple mappings in our UI very soon.

I would be very much interested in talking to you about your experience with our product. I hope that your feedback can help us to make it far better in terms of usability. Can you please send me an e-mail at zd@gooddata.com? Thanks a lot!

Agreed 100%. We are working on project backup and restore. You'll be able to backup and restore your project to selected savepoint.

We are obviously investigating the issue that you have reported earlier.

We are now recognizing the ISO format (YYYY-MM-DD). We are releasing an update this week that recognizes the US format (MM/DD/YYYY).

Regards

ZD

Davis,

You are right, we can't automatically strip any alpha characters from a number. This is something what we have just implemented and will be deploying very soon in our data cleansing pack.

Currently you unfortunately to remove all non-numeric characters like %, currency symbols (e.g. $) from your CSV. You also need to remove commas as the thousands separator.

I apologize for the inconvenience. It'll be better soon.

Regards

ZD

Alvin,

ad a) we don't have good public benchmarks of our service available. I can offer you some assistance with loading larger amounts of data and trying out your specific usecase. The performance very much depends on reports, number of rows, columns, data types etc. Generally we can handle the performance level that you are talking about.

ad b) We have some Python examples how to use our REST API to automate the data uploads. I can provide you with this example.

ad c) There are upload limits and project data size limits that apply to our community projects. These are not 'hard' technology limits. We can give you access to environment that doesn't have any limits (except the limits of underlying technology)

d) Integrating snowflake schema is easy. We have numerous toold that we use internally for migrating star schema to our project.

I'm curious about your use case. I think that we can both enjoy a short phone call. Please let me know if you are interested at zd at gooddata.com.

Thanks

ZD

We are aware of this problem. Unfortunately the product still doesn't support IE8. We plan on supporting IE8 in 10 weeks.

Thank you for feedback!

Nathan, we are aware of this problem and it is in fact fixed. The fix should be applied to our production platform in about a week.

Thanks

ZD

And it has been fixed. It will go out in the next platform release.

Yes, we know about the resizing issues. Our developers are enhancing it. You should see fixes soon.

Many thenks for your useful feedback!

ZD

You are right, currently the embedded reports are not interactive. We are working on the interactivity in this sprint (first of three weeks) and we hope that we will be able to deliver more interactive dashboards in 5 weeks from now. We are including following features:

* Predefined drill down. The author of the report can specify the way how the dashboards users can drill down (by double-clicking on attribute heading or value).
* Predefined break down. The author of the report can specify the way how the dashboards users can break down metric values (by double-clicking on metric value).
* Sorting via the up/down arrows.
* Simple filtering (exclude / show only).

Hopefully this is going to be enough for what you need.

ZD

Ufff. You are right, I was able to replicate the behavior. :( We are going to fix this one asap.

Thanks for reporting this!

Did you try to use your browser's back button? You should be able to simply undo any analysis step via the Back button.

Absolutely. We were talking about better highlighting of unsaved report revision. This will come soon. Thank you for helpful fedback!

Are you confirming your metric and attribute selections via the green "Done" button at the bottom of the What and How sections? You need to do o. I admit that we need to better enforce the explicit confirmation.

I agree that some hint regarding what to do next would be useful. Regarding the NaN issue, what browser are you using?

This is clearly bug. Nice catch!

Can you tell us more about the error you are experiencing? Is there an issue with incorrect diacritics or the upload does not work for you at all?