Electronic Sow Feeder data processing

rtniblett15 — Thu, 03 Nov 2022 02:08:54 GMT

I have a lengthy question for you SAS experts.

I am conducting an experiment using an electronic sow feeder, where animals are allowed to visit a feeder multiple times per day, thus presenting a final dataset with many visits per day for each animal. Each time feed is dispensed, if there is an allotment left, a bodyweight is recorded. I have predicted weight values based on two real weights (beginning of test and end of test) and want to remove outliers based on a +/- 5% window of this predicted weight. The weights are predicted for each day of test for each animal. I would assume that is a many to one comparison. With the same dataset, I want to sum the feed intake for each animal by day. There is more to the story, but this is where I want to start. I have attached a portion of the dataset that I am working with.

Obs desc entry exit location GiltID RFID Trt Duration Feed1 Feed2 Wt Controller Station FV NFV entry2 exit2 feed entrydate1234567891011121314

Gestation	01JAN60:12:26:45	01JAN60:12:26:45	1	220910	9.85E14	AD	873	2.000	198.414	1	2	1	0	01SEP22:10:33:36	01SEP22:10:48:00	2.000	01SEP2022
Gestation	01JAN60:12:26:46	01JAN60:12:26:46	1	220910	9.85E14	AD	9	0.183	0.000	1	2	1	0	01SEP22:18:00:00	01SEP22:18:00:00	0.183	01SEP2022
Gestation	01JAN60:12:26:46	01JAN60:12:26:46	1	220910	9.85E14	AD	1230	2.815	202.382	1	2	1	0	01SEP22:18:43:12	01SEP22:19:12:00	2.815	01SEP2022
Gestation	01JAN60:12:26:46	01JAN60:12:26:46	1	220910	9.85E14	AD	137	0.000	0.000	1	2	0	1	01SEP22:20:52:48	01SEP22:20:52:48	0.000	01SEP2022
Gestation	01JAN60:12:26:46	01JAN60:12:26:46	1	220910	9.85E14	AD	1774	3.999	208.776	1	2	1	0	02SEP22:06:28:48	02SEP22:06:57:36	3.999	02SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	37	0.000	0.000	1	2	0	1	02SEP22:18:43:12	02SEP22:18:43:12	0.000	02SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	51	0.000	0.000	1	2	0	1	02SEP22:19:12:00	02SEP22:19:12:00	0.000	02SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	143	0.000	0.000	1	2	0	1	02SEP22:19:55:12	02SEP22:19:55:12	0.000	02SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	2	0.000	0.000	1	2	0	1	02SEP22:20:09:36	02SEP22:20:09:36	0.000	02SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	215	0.366	210.539	1	2	1	0	03SEP22:04:19:12	03SEP22:04:33:36	0.366	03SEP2022
Gestation	01JAN60:12:26:47	01JAN60:12:26:47	1	220910	9.85E14	AD	2183	3.633	209.217	1	2	1	0	03SEP22:10:04:48	03SEP22:10:33:36	3.633	03SEP2022
Gestation	01JAN60:12:26:48	01JAN60:12:26:48	1	220910	9.85E14	AD	30	0.000	0.000	1	2	0	1	03SEP22:15:21:36	03SEP22:15:21:36	0.000	03SEP2022
Gestation	01JAN60:12:26:48	01JAN60:12:26:48	1	220910	9.85E14	AD	102	0.000	0.000	1	2	0	1	03SEP22:18:14:24	03SEP22:18:14:24	0.000	03SEP2022
Gestation	01JAN60:12:26:48	01JAN60:12:26:48	1	220910	9.85E14	AD	20	0.000	0.000	1	2	0	1	03SEP22:18:28:48	03SEP22:18:28:48	0.000	03SEP2022

The predicted values for this animal would be as follows:

D0	220910	231	219.45	242.55
D1	220910	233.0952	221.4405	244.75
D2	220910	235.1905	223.431	246.95
D3	220910	237.2857	225.4214	249.15

I had another dataset with a day of test column, but it is not shown here. The columns in the predicted weight table are day, visualid, -5%, PW, +5%. I am not sure if this is possible, but I can only assume the code is more advanced than my current skill level. Some of my requests can be done in a pivot table in excel, but not the outlier removal without formulation.

Re: Electronic Sow Feeder data processing

ballardw — Thu, 03 Nov 2022 17:10:09 GMT

Instructions here: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-data-AKA-generate/ta-p/258712 will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the </> icon or attached as text to show exactly what you have and that we can test code against.

Only need to show enough records to allow us to perform the actions you need.

Since your "predicted" values have no column headings or description any match with the other data is going to rely on a lot of guessing. You should also provide that data in the forum of a data step. Then identify which variables in the "predicted" data are to be used to match the collected data. I don't see a D0 in the collected data so you need to describe how that relates to the collected data.

In this case, you need to provide details as to which variables are used for what. You seem to have potentially 7 columns at least that might be used to identify groups but I am not sure which ones you intend or how.

Note that typically using a correct format for a date, time or datetime variable in a procedure like Means or Summary will allow you to group by that

topic Re: Electronic Sow Feeder data processing in SAS Programming

Electronic Sow Feeder data processing

Re: Electronic Sow Feeder data processing