Skip to content
Snippets Groups Projects
Commit 4d1faaf2 authored by Adriaens, Ines's avatar Adriaens, Ines :v_tone2:
Browse files

calculate bouts

parent 01225e30
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
behaviours.py
+ 61
− 34
View file @ 4d1faaf2
......@@ -12,13 +12,13 @@ os.chdir(r"C:\Users\adria036\OneDrive - Wageningen University & Research\iAdriae
#%% import modules
from datetime import date, datetime
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib
import seaborn as sns
from uwbfunctions import behaviour_bouts, heatmap_barn
#%matplotlib qt
......@@ -38,10 +38,10 @@ del fna, path_zones
# settings
settings = {'barn' : [60,61], # [60,61,62,70,71,72,73],
'startdate' : date(2022,10,1),
'enddate' : date(2022,10,10),
'cows' : [2658,2155], # or specific cow number
settings = {'barn' : [60], # [60,61,62,70,71,72,73],
'startdate' : date(2022,10,7),
'enddate' : date(2022,10,7),
'cows' : [907], # or specific cow number
}
# files that comply with settings
......@@ -81,33 +81,8 @@ for f in fn:
data = data.sort_values(by = ["cowid","date","t"])
data = data.reset_index(drop=1)
# correct the areas
cows = data["cowid"].drop_duplicates().reset_index(drop=1)
for cow in cows:
print(cow)
days = data.loc[(data["cowid"]==cow),["date"]].drop_duplicates()
days = days.sort_values(by = ["date"]).reset_index(drop=1)
for dd in days["date"]:
print(dd)
barnnr = int(data.loc[(data["cowid"] == cow) &
(data["date"] == dd),"barn"].mean())
myarea = area_zones.loc[(area_zones["barn"] == barnnr) | \
(area_zones["barn"].isna() == True) \
,:].reset_index(drop=1)
for j in range(0,len(myarea)):
x1,y1,x2,y2,x3,y3,x4,y4 = myarea[["x1","y1","x2","y2","x3","y3","x4","y4"]].values[j]
data.loc[((data["cowid"] == cow) & (data["date"] == dd) & \
(data["xnew"] >= x1) & (data["xnew"] <= x2) & \
(data["xnew"] >= x3) & (data["xnew"] <= x4) & \
(data["ynew"] >= y1) & (data["ynew"] >= y2) & \
(data["ynew"] <= y3) & (data["ynew"] <= y4)),"area"] = myarea["area"][j]
data.loc[((data["cowid"] == cow) & (data["date"] == dd) & \
(data["xnew"] >= x1) & (data["xnew"] <= x2) & \
(data["xnew"] >= x3) & (data["xnew"] <= x4) & \
(data["ynew"] >= y1) & (data["ynew"] >= y2) & \
(data["ynew"] <= y3) & (data["ynew"] <= y4)),"zone"] = myarea["zone"][j]
del x1,y1,x2,y2,x3,y3,x4,y4
del cows, dd, days
data.loc[data["zone"].isna(),"zone"] = 8
data.loc[data["zone"] == 8 ,"area"] = "unknown"
......@@ -117,8 +92,60 @@ data["date"] = data["date"].dt.date
print(data[["cowid","barn"]].drop_duplicates())
#%% RESTING BEHAVIOUR
#%% calculate and summarize behaviours
cows = data["cowid"].drop_duplicates().reset_index(drop=1)
bouts = pd.DataFrame([])
bsums = pd.DataFrame([])
for cow in cows:
subset =
print(cow)
# ------------------------ feeding behaviour ------------------------------
interval = 5*60 # min number of seconds between separate bouts
behaviour = "feed" # feeding behaviour
min_length = 30 # do not count if shorter than 30 seconds
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour,min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
# ------------------------ drinking behaviour -----------------------------
interval = 5*60 # min number of seconds between separate bouts
behaviour = "drink" # drinking behaviour
min_length = 5 # do not count if shorter than 2 seconds
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour, min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
# ------------------------ resting behaviour ------------------------------
interval = 2*60 # min number of seconds between separate bouts
behaviour = "cubicle" # resting behaviour
min_length = 30 # do not count if shorter than 30 seconds
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour, min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
# ------------------------ concentrate feeder -----------------------------
interval = 5*60 # min number of seconds between separate bouts
behaviour = "concentrate" # concentrate feeding behaviour
min_length = 0 # always count
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour, min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
# ------------------------ waiting area ------------------------------
interval = 20*60 # min number of seconds between separate bouts
behaviour = "wait" # in the waiting area
min_length = 0 # always count
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour, min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
# ------------------------ unknown area ------------------------------
interval = 5*60 # min number of seconds between separate bouts
behaviour = "unknown" # unknown behaviour
min_length = 10 # do not count if shorter than 10 seconds
df,summary = behaviour_bouts(data.loc[data["cowid"] == cow], interval, behaviour, min_length)
bouts = pd.concat([bouts,df],axis = 0).reset_index(drop=1)
bsums = pd.concat([bsums,summary],axis = 0).reset_index(drop=1)
bin.py 0 → 100644
+ 29
− 0
View file @ 4d1faaf2
"""
# correct the areas
cows = data["cowid"].drop_duplicates().reset_index(drop=1)
for cow in cows:
print(cow)
days = data.loc[(data["cowid"]==cow),["date"]].drop_duplicates()
days = days.sort_values(by = ["date"]).reset_index(drop=1)
for dd in days["date"]:
print(dd)
barnnr = int(data.loc[(data["cowid"] == cow) &
(data["date"] == dd),"barn"].mean())
myarea = area_zones.loc[(area_zones["barn"] == barnnr) | \
(area_zones["barn"].isna() == True) \
,:].reset_index(drop=1)
for j in range(0,len(myarea)):
x1,y1,x2,y2,x3,y3,x4,y4 = myarea[["x1","y1","x2","y2","x3","y3","x4","y4"]].values[j]
data.loc[((data["cowid"] == cow) & (data["date"] == dd) & \
(data["xnew"] >= x1) & (data["xnew"] <= x2) & \
(data["xnew"] >= x3) & (data["xnew"] <= x4) & \
(data["ynew"] >= y1) & (data["ynew"] >= y2) & \
(data["ynew"] <= y3) & (data["ynew"] <= y4)),"area"] = myarea["area"][j]
data.loc[((data["cowid"] == cow) & (data["date"] == dd) & \
(data["xnew"] >= x1) & (data["xnew"] <= x2) & \
(data["xnew"] >= x3) & (data["xnew"] <= x4) & \
(data["ynew"] >= y1) & (data["ynew"] >= y2) & \
(data["ynew"] <= y3) & (data["ynew"] <= y4)),"zone"] = myarea["zone"][j]
del x1,y1,x2,y2,x3,y3,x4,y4
del cows, dd, days
"""
\ No newline at end of file
heatmap_visualisation.py → uwbfunctions.py
+ 101
− 0
View file @ 4d1faaf2
......@@ -123,6 +123,107 @@ def heatmap_barn(data_x,data_y,xstep,ystep,x_lim,y_lim):
return ax
#%% bout calculations
def behaviour_bouts(data, interval, behaviour,min_length):
"""
Parameters
----------
data : Pandas DataFrame
data array with at least:
- t : time stamp
- xnew, ynew : preprocessed position
- area : area (walking, drink_trough, cubicle_A-B-C, feed_rack)
- zone
interval : numeric
time in seconds that the behaviour is discontinued to belong to another
bout
Returns
-------
df : Pandas DataFrame
dataframe with bouts of selected behaviour, according to max interval
of interval between two bouts
summary : Pandas DataFrame
summary of the behaviour for that day
"""
#---------------------------- for development -----------------------------
# interval = 5*60
# behaviour = "unknown"
# min_length = 5
# sep = {"cubicle": 60, "walking" : 3*60, "drink_trough" : 5*60}
#---------------------------- for development -----------------------------
# select data of behaviour of interest and calculate discontinuities
subset = data.loc[(data["area"].str.contains(behaviour)),:].copy().reset_index(drop=1)
if len(subset) > 0:
subset["gap"] = subset["t"].diff()
subset["gapprev"] = np.nan
subset.iloc[0:-1,subset.columns == "gapprev"] = subset.iloc[1:,subset.columns == "gap"]
# keep where gap > interval and add previous without gap to mark end of bout
idx = np.array(subset.loc[subset["gap"] > interval].index.values)
idxstart = np.sort(np.append(idx,subset.index.values[0]))
idx = np.array(subset.loc[subset["gapprev"] > interval].index.values)
idxend = np.sort(np.append(idx,subset.index.values[-1]))
df = subset.loc[idxstart,:].reset_index(drop=1)
df["end_time"] = subset.loc[idxend,"t"].values
# add time to check with figures
df["bhour"] = np.floor(df["t"]/3600)
df["bmin"] = np.floor((df["t"]-df["bhour"]*3600)/60)
df["ehour"] = np.floor(df["end_time"]/3600)
df["emin"] = np.floor((df["end_time"]-df["ehour"]*3600)/60)
# for the bout length, calculate both true number of times behaviour
# is recorded, and end - begin incl the other behaviours in between
# => if "unknown", also put it as the behaviour of interest
df["len1"] = round((df["end_time"] - df["t"])/60,1)
# calculate true number of seconds this behaviour is noted
df["len2"] = np.nan
for i in range(0,len(df)):
df.loc[i,"len2"] = round(len(data.loc[(data["t"] >= df["t"][i]) & \
(data["t"] <= df["end_time"][i]) & \
((data["area"].str.contains(behaviour)) | \
(data["area"].str.contains("unknown")))])/60,1)
# only keep bouts > min_length
df = df.loc[df["len1"] > min_length,:].reset_index(drop=1)
# calculate gap between successive bouts
df["gap"] = np.nan
df.iloc[1:,df.columns == "gap"] = np.round((df["t"][1:].values - df["end_time"][0:-1].values)/60)
# add behaviour and select data
df["behaviour"] = behaviour
df["btime"] = pd.to_datetime(df["date"]) + pd.to_timedelta(df["t"],unit = 's')
df["etime"] = pd.to_datetime(df["date"]) + pd.to_timedelta(df["end_time"],unit = 's')
df = df[["cowid","barn","behaviour","date","t","end_time","btime","etime",
"gap","len1","len2"]]
df = df.rename(columns = {"end_time" : "end","t" : "start"})
# calculate summary
summary = df[["cowid","barn","date","behaviour","len1","len2"]].groupby(by = ["cowid","barn","date","behaviour"]).mean().round(1).reset_index()
summary = summary.rename(columns = {"len1" : "len1_av", "len2" : "len2_av"})
summary["no_bouts"] = len(df)
summary["gap_med"] = df["gap"].median()
summary["total1"] = df["len1"].sum()
summary["total2"] = df["len2"].sum()
#print(summary)
else:
df = pd.DataFrame([],columns = ["cowid","barn","behaviour","date","start",
"end","btime","etime",
"gap","len1","len2"])
summary = pd.DataFrame([],columns = ["cowid","barn","date","behaviour",
"len1_av","len2_av","no_bouts",
"gap_med","total1","total2"])
return df,summary
"""
#%% usage
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment