separated in bouts even when the same behavior is repeated?

jcolomb / HCS_analysis

This is an open science project, here the code to analyse data obtained from mice home cage monitoring (especially homecagescan,HCS, cleversys Inc.)

MIT License

4 stars 2 forks source link

separated in bouts even when the same behavior is repeated? #9

Open jcolomb opened 6 years ago

jcolomb commented 6 years ago

To get exact timing around the change in light conditions, one would need to recompute the minute data from the raw data. To do this we need to know:

To do other calculation (sequence analysis for example), it would be good to know why some behavior are recorded as long duration, while others are separated in bouts even when the same behavior is repeated. (why we can have "comedown .24 -- comedown .24" instead of "comedown .48")

jcolomb commented 6 years ago

Answer: Two consecutive behaviors should typically NOT be the same behavior. It should be merged. Only case where you may see this is, if certain behaviors are "Ignored" during analysis. So, for example, if you had Rear UP followed by Come Down followed by Rear Up again, but, for some reason you ignored Come Down, then, you may see 2 Rear Up bouts juxtaposed. Normally, this should not be the case. Let me know if you see such examples.

jcolomb commented 6 years ago

What I was refering to is the behavior sequence exported in a excel file. As an example, see: see https://github.com/jcolomb/HCS_analysis/blob/master/data/Ro_testdata/HCS3_output/150608_Testdata_HCS_groupA/150608_Testdata_HCS_beh_20.xlsx

line 195 and 196 depict a turn of 24 ''.

There are probably a ton of such examples...

jcolomb commented 6 years ago

got it in a mbr file too:

817 823 2 823 829 2 829 835 2

it seems to be problematic only for turns, but I did not test it.