Eeeeek!!! Your stored process failed. It worked when in EG – how could this disaster have happened? If you check the log, you may see an error similar to the following about the Libname not being assigned. The library where you stored the data, in this case, MYLIB was either available to SAS EG but not widely known to the SAS BI system or maybe it has a different name in the metadata. To move forward, you need to add a LIBNAME statement that describes where the metadata is located to the stored process. So you would add the following to the code: LIBNAME name meta liburi=”SASLIBRARY?@name=’metadata_name‘ “; You have seen LIBNAME statements before – but wow this one has some extra goodies. My Enchanted Mouse Tricks You may be wondering how I knew what to assign the values – clearly it must be some kind of BI white magic. Yes, using my enchanted mouse in EG, I pointed-and-clicked my way to the answer. In the Server List, I located the dataset being used (SALES_CANDY_HISTORY) and noted the library name, which is Candy. If you review the Candy library properties, you can see some information about the library, where it is stored […]
Example 9.7: New stuff in SAS 9.3– Frailty models
Shared frailty models are a way of allowing correlated observations into proportional hazards models. Briefly, instead of l_i(t) = l_0(t)e^(x_iB), we allow l_ij(t) = l_0(t)e^(x_ijB + g_i), where observations j are in clusters i, g_i is typically norma…
Example 9.7: New stuff in SAS 9.3– Frailty models
Shared frailty models are a way of allowing correlated observations into proportional hazards models. Briefly, instead of l_i(t) = l_0(t)e^(x_iB), we allow l_ij(t) = l_0(t)e^(x_ijB + g_i), where observations j are in clusters i, g_i is typically norma…
Example 9.7: New stuff in SAS 9.3– Frailty models
Shared frailty models are a way of allowing correlated observations into proportional hazards models. Briefly, instead of l_i(t) = l_0(t)e^(x_iB), we allow l_ij(t) = l_0(t)e^(x_ijB + g_i), where observations j are in clusters i, g_i is typically norma…
Web Report Studio: Who stole my pie (chart)?
Ok – I”ll just say it. I hate pie charts. If there are more than a few categories – then the user has to interpret the ranking order, it slows them down, and they may loose interest in the report that you so lovingly prepared. Maybe you say – “Tricia, add the numbers! Geez – you’re so dramatic!” Then I will calmly reply to your dramatic outburst, if I want to have numbers – then why not just provide a data table? Let’s look at some examples so I can use the power of screen captures to sway you to my way of thinking. 30-Second Test Consider the following beautiful pie chart from Web Report Studio – you have 30 seconds to decide the product had the largest sales. It will take 30 seconds because you have to determine which slice actually is the largest (and there are two competitors) and then correlate the slice color to the product name in the legend. Go ahead, I’ll wait. Ok – let’s try the method of adding the numbers. Note that this beauty contest winner is getting cluttered. To make room for the numbers, the chart is getting smaller. I could use percentages to make the […]
The default settings are not good enough
Today, I came across this article that discusses how software users rarely change the settings in the applications that they use. Users assume that the software vendor sets the default values as they are for a reason, and who are they to set them otherwise? I’m a software developer (and […]