def pearson ( x, y):
p = [x[i]*y[i] for i in range(len(y))]
if sd(x) != 0:
return(
( mean(p) - mean(x)*mean(y) ) / ( sd(x)* sd(y) )
)
else:
return( 0.0)
def wbb(self, i):
#domain=self.prices[i-self.ma_n:]
#domainsd=sd(domain)
#if i in self.sddic:
# sdval=self.sddic[i]
#else:
# sdval=sd(domain)
# self.sddic[i]=sdval
if i in self.madic:
ma = self.madic[i]
else:
ma = self.ma(self.ma_n, i)
self.madic[i] = ma
print("MOVING AVERAGES: ", ma)
return (ma[-1] + sd(ma)) - (ma[-1] - sd(ma))
def normalize_column(data: DataFrame, column: str):
"""
Find the Z score equivalent of each column
"""
m = mean(data[column])
s = sd(data[column])
return data[column].map(lambda x: (x - m) / s)
def getSummary(self, e=None):
string = ""
#string = "Demo: " +self.lastDemoName +" Map: " + Demo_GetWorld(self.lastDemoName) +"\r\n"; #TODO
if e != None:
string += "Test number: " + str(len(self.fps)) + "\r\n"
string += "FPS: " + str(e.fps) + "\r\n"
string += "Frames Drew: " + str(e.frames) + "\r\n"
string += "Run Time: " + str(e.time) + "\r\n"
string += "Number of tests: " + str(len(self.fps)) + "\r\n"
string += "Average FPS: " + str(statistics.mean(self.fps)) + "\r\n"
string += "Std Dev FPS: " + str(statistics.sd(self.fps)) + "\r\n"
return string
#from statistics import * from statistics import stdev as sd, mean as m example_list = [3, 5, 7, 8, 23, 2, 5, 3, 51, 52, 626, 62, 623] x = sd(example_list) print(x) x = m(example_list) print(x) ''' import statistics as s example_list = [3,5,7,8,23,2,5,3,51,52,626,62,623] x = s.mean(example_list) print(x) x = s.median(example_list) print(x) x = s.stdev(example_list) print(x) x = s.variance(example_list) print(x) '''
exList = [5,2,3,7,8,9,4,5,6,1,0,3,5] import statistics as s print(s.mean(exList)) print(s.median(exList)) print(s.mode(exList)) print(s.stdev(exList)) print(s.variance(exList)) from statistics import mean print(mean(exList)) from statistics import mean as m print(m(exList)) from statistics import mean, stdev print(mean(exList), stdev(exList)) from statistics import mean as m, stdev as sd print(m(exList), sd(exList)) from statistics import * print(mean(exList), stdev(exList))
list_performance.append((v, e, end_time - start_time))
gg.append(end_time - start_time)
end_time_big = time.time()
for row in list_performance:
sheet.append(row)
filepath = "/content/drive/My Drive/301_Project/" + str(v) + '_' + str(i)
wb.save(filepath)
t_val = t.ppf(0.95, i)
mean = 0
for i in range(len(gg)):
mean += gg[i]
mean = mean / len(gg)
st_dev = sd(gg)
std_mean_err = st_dev / i**(0.5)
low = mean - t_val * std_mean_err
high = mean + t_val * std_mean_err
print("#vertex: ", n)
print("#try:", i)
print('mean: ', mean)
print(low, high)
print('time:', end_time_big - start_time_big)
from scipy.stats import t
tt = 0
fasta[key] += line.strip()
return fasta
#iterate over fastas in a directory
d = 'LinkerFastasNew'
gl_means = []
gl_sds = []
gl_gap_means = []
gl_gap_sds = []
for fn in ls(d):
fasta = readFasta(open(d + '/' + fn, 'r'))
lengths = []
gaps = []
for entry in fasta:
lengths.append(len(fasta[entry]))
gaps.append(fasta[entry].count('-'))
gl_means.append(str(sum(lengths) / len(lengths)) + '\n')
gl_sds.append(str(sd(lengths)) + '\n')
gl_gap_means.append(str(sum(gaps) / len(lengths)) + '\n')
gl_gap_sds.append(str(sd(gaps)) + '\n')
out_gl_means = open('Stats/' + d + '_means.txt', 'w')
out_gl_means.writelines(gl_means)
out_gl_means = open('Stats/' + d + '_sds.txt', 'w')
out_gl_means.writelines(gl_sds)
out_gl_means = open('Stats/' + d + '_gap_means.txt', 'w')
out_gl_means.writelines(gl_gap_means)
out_gl_means = open('Stats/' + d + '_gap_sds.txt', 'w')
out_gl_means.writelines(gl_gap_sds)
def feature_list(user_id: str, session: str, tap_feature: str, task_name: str,
window: DataFrame):
"""
A list of features identifying a tap
:param window: a DataFrame with just the data needed for the computation
"""
if window.shape[0] == 0:
return None
#Add user ID, session, task name
features = [user_id, session, task_name]
#Add orientation
orientation = mode(window['Phone_orientation_accel'])
features.append(orientation)
#Add tap type
features.append(tap_feature)
lead_file = 'Accelerometer.csv'
time_col = x_columns[lead_file]
before_start = window[window[tap_feature] == 4].index[0]
during_start = window[window[tap_feature] == 2].index[0]
after_start = window[window[tap_feature] == 3].index[0] + 1
after_end = window[window[tap_feature] == 5].index[0]
before = window.loc[before_start:during_start]
during = window.loc[during_start:after_start]
after = window.loc[after_start:after_end + 1]
if during.shape[0] < 2:
# If there were none or one measurements during the tap,
# add the closest ones
during = window[during_start - 1:after_start + 1]
for file_name in file_names:
for y in y_columns[file_name]:
# Feature 1: Mean during
mean_during = mean(during[y])
# Feature 2: SD during
sd_during = sd(during[y])
# Feature 3: Difference before/after
mean_before = mean(before[y])
mean_after = mean(after[y])
difference_before_after = mean_after - mean_before
# Feature 4: Net change from tap
net_change_due_to_tap = mean_during - mean_before
# Feature 5: Maximal change from tap
max_tap = max(during[y])
max_change = max_tap - mean_before
# Feature 6: Restoration time
avgDiffs = []
for j in range(after[y].shape[0]):
subsequentValues = after[y].iloc[j:]
subsequentDistances = subsequentValues.map(
lambda x: abs(x - mean_before))
averageDistance = mean(subsequentDistances)
avgDiffs.append(averageDistance)
time_of_earliest_restoration = min(avgDiffs)
restoration_time = time_of_earliest_restoration - during[
time_col].iloc[-1]
# Feature 7: Normalized duration
t_before_center = (before[time_col].iloc[0] +
before[time_col].iloc[-1]) / 2
t_after_center = (after[time_col].iloc[0] +
after[time_col].iloc[-1]) / 2
normalized_duration = (t_after_center - t_before_center) / (
mean_after - mean_before)
# Feature 8: Ndormalized duration max
t_max_in_tap = during[during[y] == max_tap][time_col].iloc[0]
normalized_duration_max = (t_after_center -
t_max_in_tap) / (mean_after - max_tap)
features += [
mean_during, sd_during, difference_before_after,
net_change_due_to_tap, max_change, restoration_time,
normalized_duration, normalized_duration_max
]
if random.choice(range(100)) == 0:
plot_tap('Plots/Project/' + session, before, during, after, time_col)
return features