def test_hc_param():
# Check get_optimal_hc_params returns appropriate parameters
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
method, dist = clustering.get_optimal_hc_params(preped_data)
assert method in ['ward', 'average', 'complete']
assert dist in ['cityblock', 'euclidean', 'chebychev']
def total_amount(strain, mouse, day, feature):
"""Returns the total amount consumed/moved for one of the following features:
food ("F"), water ("W"), and locomotion ("L").
Parameters
----------
strain : int
mouse : int
day : int
feature : {'F', 'W', 'L'}
The feature used. 'F' for food, 'W' for water, 'L' for locomotion.
Returns
-------
The total amount consumed/moved for the given feature in the mouse-day.
"""
df = data.load_all_features()
df = _select_strain_mouse_day_in_data_frame(df, strain, mouse, day)
if df.shape[0] == 0:
raise ValueError(
"No consumption data available for specified mouse-day.")
feature_names = {'W': 'Water', 'F': 'Food', 'L': 'Distance'}
if feature not in feature_names.keys():
raise ValueError("Feature must be one of :" +
str(set(feature_names.keys())))
return df[feature_names[feature]].sum()
def total_amount(strain, mouse, day, feature):
"""Returns the total amount consumed/moved for one of the following features:
food ("F"), water ("W"), and locomotion ("L").
Parameters
----------
strain : int
mouse : int
day : int
feature : {'F', 'W', 'L'}
The feature used. 'F' for food, 'W' for water, 'L' for locomotion.
Returns
-------
The total amount consumed/moved for the given feature in the mouse-day.
"""
df = data.load_all_features()
df = _select_strain_mouse_day_in_data_frame(df, strain, mouse, day)
if df.shape[0] == 0:
raise ValueError(
"No consumption data available for specified mouse-day.")
feature_names = {'W': 'Water', 'F': 'Food', 'L': 'Distance'}
if feature not in feature_names.keys():
raise ValueError("Feature must be one of :" + str(set(feature_names
.keys())))
return df[feature_names[feature]].sum()
def test_hc_param():
# Check get_optimal_hc_params returns appropriate parameters
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
method, dist = clustering.get_optimal_hc_params(preped_data)
assert method in ['ward', 'average', 'complete']
assert dist in ['cityblock', 'euclidean', 'chebychev']
def test_fit_kmeans():
# Check get_optimal_fit_kmeans returns expected result
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
mouse_day_X = preped_data[:, 2:]
res = clustering.get_optimal_fit_kmeans(
mouse_day_X, num_clusters=range(2, 17), raw=False)
assert len(res) == 2
assert len(res[0]) == 15
assert len(res[1][0]) == 170
assert len(set(res[1][14])) <= 16
# silhouette score should be between -1 and 1
assert all(value < 1 for value in res[0])
assert all(value > -1 for value in res[0])
def test_fit_hc():
# Check fit_hc returns appropriate result
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
mouse_day_X = preped_data[:, 2:]
res = clustering.fit_hc(mouse_day_X, "average", "chebychev",
num_clusters=range(2, 17))
assert len(res) == 2
assert len(res[0]) == 15
assert len(res[1][0]) == 170
assert len(set(res[1][14])) <= 16
# silhouette score should be between -1 and 1
assert all(value < 1 for value in res[0])
assert all(value > -1 for value in res[0])
def test_fit_kmeans():
# Check get_optimal_fit_kmeans returns expected result
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
mouse_day_X = preped_data[:, 2:]
res = clustering.get_optimal_fit_kmeans(mouse_day_X,
num_clusters=range(2, 17),
raw=False)
assert len(res) == 2
assert len(res[0]) == 15
assert len(res[1][0]) == 170
assert len(set(res[1][14])) <= 16
# silhouette score should be between -1 and 1
assert all(value < 1 for value in res[0])
assert all(value > -1 for value in res[0])
def test_fit_hc():
# Check fit_hc returns appropriate result
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
mouse_day_X = preped_data[:, 2:]
res = clustering.fit_hc(mouse_day_X,
"average",
"chebychev",
num_clusters=range(2, 17))
assert len(res) == 2
assert len(res[0]) == 15
assert len(res[1][0]) == 170
assert len(set(res[1][14])) <= 16
# silhouette score should be between -1 and 1
assert all(value < 1 for value in res[0])
assert all(value > -1 for value in res[0])
def test_prep_data():
# Check prep_data return the correct dimension
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
assert preped_data.shape == (170, 20)
def test_all_features_loader():
# Checking load_all_features returns a data frame of the correct dimension
all_features = data.load_all_features()
assert all_features.shape == (21131, 13)
def test_intervals():
# This is a place holder. Not sure this is correct.
all_features = data.load_all_features()
assert Intervals(all_features).measure() == 11.0
from mousestyles import data
from mousestyles.classification import clustering
from mousestyles.visualization import plot_clustering
# load data
mouse_data = data.load_all_features()
# mouse inidividual
mouse_dayavgstd_rsl = clustering.prep_data(mouse_data, melted=False, std = True, rescale = True)
# optimal parameters
method, dist = clustering.get_optimal_hc_params(mouse_day=mouse_dayavgstd_rsl)
# fit hc
sils_hc, labels_hc = clustering.fit_hc(
mouse_day_X=mouse_dayavgstd_rsl[:,2:],
method=method, dist=dist, num_clusters=range(2,17))
# plot and get the distance matrxix
Z = plot_clustering.plot_dendrogram(
mouse_day=mouse_dayavgstd_rsl, method=method, dist=dist)
def test_prep_data():
# Check prep_data return the correct dimension
mouse_data = data.load_all_features()
preped_data = clustering.prep_data(mouse_data)
assert preped_data.shape == (170, 20)
def test_all_features_loader():
all_features = data.load_all_features()
assert_equal(all_features.shape, (21131, 13))
from mousestyles import data
from mousestyles.classification import clustering
from mousestyles.visualization import plot_clustering
# load data
mouse_data = data.load_all_features()
# mouse inidividual
mouse_dayavgstd_rsl = clustering.prep_data(mouse_data,
melted=False,
std=True,
rescale=True)
# optimal parameters
method, dist = clustering.get_optimal_hc_params(mouse_day=mouse_dayavgstd_rsl)
# fit hc
sils_hc, labels_hc = clustering.fit_hc(mouse_day_X=mouse_dayavgstd_rsl[:, 2:],
method=method,
dist=dist,
num_clusters=range(2, 17))
# result
result = clustering.cluster_in_strain(mouse_dayavgstd_rsl[:, 0], labels_hc[4])
# plot
plot_clustering.plot_strain_cluster(count_data=result, groupby_cluster=True)
def aggregate_interval(strain, mouse, feature, bin_width):
"""
Aggregate the interval data based on n-minute time
intervals, return a time series.
Parameters
----------
strain: int
nonnegative integer indicating the strain number
mouse: int
nonnegative integer indicating the mouse number
feature: {"AS", "F", "M_AS", "M_IS", "W"}
"AS": Active state probalibity
"F": Food consumed (g)
"M_AS": Movement outside homebase
"M_IS": Movement inside homebase
"W": Water consumed (g)
bin_width: number of minutes of time interval for data aggregation
Returns
-------
ts: pandas.tseries
a pandas time series of length 12(day)*24(hour)*60(minute)/n
"""
# Input Check
if (not isinstance(strain, int)) or (strain < 0):
raise ValueError('Strain must be a non-negative integer')
if (not isinstance(mouse, int)) or (mouse < 0):
raise ValueError('Mouse value must be a non-negative integer')
if feature not in INTERVAL_FEATURES:
raise ValueError(
'Input value must in {"AS", "F", "M_AS", "M_IS", "W"}')
if (not isinstance(bin_width, int)) or bin_width < 0 or bin_width > 1440:
raise ValueError(
'Bin width (minutes) must be a non-negative integer below 1440')
# load data
intervals = data.load_intervals(feature)
mouse_data = intervals.loc[(intervals['strain'] == strain)
& (intervals['mouse'] == mouse)]
# build data frame
days = sorted(np.unique(mouse_data['day']))
bin_count = int(24 * 60 / bin_width)
time_behaviour = np.repeat(0.0, bin_count * len(days))
bin_length = bin_width * 60
for j in days:
df = mouse_data.loc[mouse_data['day'] == j]
start_end = data.load_start_time_end_time(strain, mouse, j)
start = np.asarray(df['start']) - start_end[0]
end = np.asarray(df['stop']) - start_end[0]
for i in range(len(start)):
start_time = start[i]
end_time = end[i]
start_index = int(start_time / (bin_width * 60))
end_index = int(end_time / (bin_width * 60))
if start_index == end_index:
time_behaviour[start_index +
j * bin_count] += end_time - start_time
elif end_index - start_index == 1:
time_behaviour[
start_index +
j * bin_count] += bin_length * end_index - start_time
time_behaviour[end_index +
j * bin_count] += end_time % bin_length
else:
time_behaviour[start_index +
j * bin_count] += bin_length * (start_index +
1) - start_time
time_behaviour[end_index +
j * bin_count] += end_time % bin_length
time_behaviour[start_index + j * bin_count + 1:end_index +
j * bin_count] += bin_length
if feature == 'F' or feature == 'W':
all_feature = data.load_all_features()
group = all_feature[[
"strain", "mouse", "day", "hour", "Food", "Water"
]].groupby(["strain", "mouse", "day"]).sum()
group = group.reset_index()
mouse_data = group.loc[(group['strain'] == strain)
& (group['mouse'] == mouse)].copy()
mouse_data.loc[:, 'day'] = np.arange(len(mouse_data))
for i in mouse_data['day'].astype('int'):
if feature == 'F':
food_amount = float(mouse_data['Food'][mouse_data['day'] == i])
time_behaviour[(bin_count * i):(bin_count * (i + 1))] /= sum(
time_behaviour[(bin_count * i):(bin_count * (i + 1))])
time_behaviour[(bin_count * i):(bin_count *
(i + 1))] *= food_amount
else:
food_amount = float(
mouse_data['Water'][mouse_data['day'] == i])
time_behaviour[(bin_count * i):(bin_count * (i + 1))] /= sum(
time_behaviour[(bin_count * i):(bin_count * (i + 1))])
time_behaviour[(bin_count * i):(bin_count *
(i + 1))] *= food_amount
if feature == 'AS':
time_behaviour /= (bin_width * 60)
ts = pd.Series(time_behaviour,
index=pd.date_range('01/01/2014',
periods=len(time_behaviour),
freq=str(bin_width) + 'min'))
return (ts)
def aggregate_interval(strain, mouse, feature, bin_width):
"""
Aggregate the interval data based on n-minute time
intervals, return a time series.
Parameters
----------
strain: int
nonnegative integer indicating the strain number
mouse: int
nonnegative integer indicating the mouse number
feature: {"AS", "F", "M_AS", "M_IS", "W"}
"AS": Active state probalibity
"F": Food consumed (g)
"M_AS": Movement outside homebase
"M_IS": Movement inside homebase
"W": Water consumed (g)
bin_width: number of minutes of time interval for data aggregation
Returns
-------
ts: pandas.tseries
a pandas time series of length 12(day)*24(hour)*60(minute)/n
"""
# Input Check
if (not isinstance(strain, int)) or (strain < 0):
raise ValueError(
'Strain must be a non-negative integer')
if (not isinstance(mouse, int)) or (mouse < 0):
raise ValueError(
'Mouse value must be a non-negative integer')
if feature not in INTERVAL_FEATURES:
raise ValueError(
'Input value must in {"AS", "F", "M_AS", "M_IS", "W"}')
if (not isinstance(bin_width, int)) or bin_width < 0 or bin_width > 1440:
raise ValueError(
'Bin width (minutes) must be a non-negative integer below 1440')
# load data
intervals = data.load_intervals(feature)
mouse_data = intervals.loc[
(intervals['strain'] == strain) & (intervals['mouse'] == mouse)]
# build data frame
days = sorted(np.unique(mouse_data['day']))
bin_count = int(24 * 60 / bin_width)
time_behaviour = np.repeat(0.0, bin_count * len(days))
bin_length = bin_width * 60
for j in days:
df = mouse_data.loc[mouse_data['day'] == j]
start_end = data.load_start_time_end_time(strain, mouse, j)
start = np.asarray(df['start']) - start_end[0]
end = np.asarray(df['stop']) - start_end[0]
for i in range(len(start)):
start_time = start[i]
end_time = end[i]
start_index = int(start_time / (bin_width * 60))
end_index = int(end_time / (bin_width * 60))
if start_index == end_index:
time_behaviour[start_index + j *
bin_count] += end_time - start_time
elif end_index - start_index == 1:
time_behaviour[
start_index + j *
bin_count] += bin_length * end_index - start_time
time_behaviour[end_index + j *
bin_count] += end_time % bin_length
else:
time_behaviour[
start_index + j *
bin_count] += bin_length * (start_index + 1) - start_time
time_behaviour[end_index + j *
bin_count] += end_time % bin_length
time_behaviour[start_index + j * bin_count +
1:end_index + j * bin_count] += bin_length
if feature == 'F' or feature == 'W':
all_feature = data.load_all_features()
group = all_feature[
["strain", "mouse", "day", "hour", "Food", "Water"]].groupby(
["strain", "mouse", "day"]).sum()
group = group.reset_index()
mouse_data = group.loc[(group['strain'] == strain) &
(group['mouse'] == mouse)].copy()
mouse_data.loc[:, 'day'] = np.arange(len(mouse_data))
for i in mouse_data['day'].astype('int'):
if feature == 'F':
food_amount = float(mouse_data['Food'][mouse_data['day'] == i])
time_behaviour[
(bin_count * i):(bin_count * (i + 1))] /= sum(
time_behaviour[(bin_count * i):(bin_count * (i + 1))])
time_behaviour[(bin_count * i):(bin_count *
(i + 1))] *= food_amount
else:
food_amount = float(mouse_data['Water'][
mouse_data['day'] == i])
time_behaviour[
(bin_count * i):(bin_count * (i + 1))] /= sum(
time_behaviour[(bin_count * i):(bin_count * (i + 1))])
time_behaviour[(bin_count * i):(bin_count *
(i + 1))] *= food_amount
if feature == 'AS':
time_behaviour /= (bin_width * 60)
ts = pd.Series(time_behaviour, index=pd.date_range(
'01/01/2014', periods=len(time_behaviour),
freq=str(bin_width) + 'min'))
return ts