def main(): all_recs = cyb_records.Stats() all_recs.ReadRecords() print 'Number of total stats', len(all_recs.records) cdf = CdfPerDay(all_recs.records) myplot.Cdfs(cdf) myplot.Show(title="CDF: daily usage of machines at the YMCA", xlabel = 'Distance (in m / day)', ylabel = 'Percentile')
def main():
all_recs = cyb_records.Stats()
all_recs.ReadRecords()
print 'Number of total stats', len(all_recs.records)
cdf = CdfPerMachine(all_recs.records)
myplot.Cdfs(cdf)
myplot.Show(title="CDF of cardio machine average distances",
xlabel='Average Distances',
ylabel='Probability')
def main():
all_recs = cyb_records.Stats()
all_recs.ReadRecords()
print 'Number of total stats', len(all_recs.records)
cdf = CdfPerDay_ScipyStyle(all_recs.records)
print "Fri-norm ", str(kstest(cdf[5], 'norm'))
print "Fri-Mon ", str(ks_2samp(cdf[5], cdf[1]))
print "Fri-Tue ", str(ks_2samp(cdf[5], cdf[2]))
print "Fri-Wed ", str(ks_2samp(cdf[5], cdf[3]))
print "Fri-Thu ", str(ks_2samp(cdf[5], cdf[4]))
print "Fri-Sat ", str(ks_2samp(cdf[5], cdf[6]))
print "Fri-Sun ", str(ks_2samp(cdf[5], cdf[0]))
def main():
evs = cyb_records.Stats() #"Stat", or "Events", etc.
evs.ReadRecords()
machines_events_list = split_up_machine_events(evs.records)
check_install_date_consistency(machines_events_list)
max_latency = 60 # seconds?
# testing code- will end up in another function
#percent, ave_latent = check_created_recieved_discrepancies(machines_events_list[0], max_latency)
#print ("Machine " + str(machines_events_list[0][0].sn) + " was " + str(percent) +" percent latent (cuttoff at "
# + str(max_latency) + " seconds latent) and had an average latency of " + str(ave_latent) + " seconds")
machine_entries_length = []
for machine in machines_events_list:
machine_entries_length.append((len(machine), machine))
machine_entries_length = sorted(machine_entries_length)
for machine in machine_entries_length:
print(
str(machine[1][0].sn) + " is in the stats database " +
str(machine[0]) + " times.")
def main():
# so that you don't have to type in passwords and usernames every time you connect- since you connect separately
# for each table.
myusername = raw_input("Please enter your username: "******"Please enter your password: "******"Stat", or "Events", etc.
evs.ReadRecords(username=myusername, password=mypassword)
machines_events_list = prep.split_up_machine_events(evs.records)
current_machines = [
] # list of all the attributes we are interested in for each machine, for the most current entry.
# format for a machine in current_machines:
# [sn(str), up_time(int), site_code(int), number of errors]
# grab sn, up time from the stats table, add them to the current_machines list
for machine in machines_events_list:
new_list = []
if "" in str(machine[0].sn
): # optional filtering of which records are collected.
record = latest_record(machine)
#print (str(record.sn) + str(record.up_time))#str(int((record.up_time/1000000.0))))
new_list.append(str(record.sn))
new_list.append(int(record.up_time))
current_machines.append(new_list)
# add in the facility id from the machines table.
evs = cyb_records.Machines()
evs.ReadRecords(username=myusername, password=mypassword)
for record in evs.records:
current_sn = str(record.sn)
for machine in current_machines:
machine_sn = machine[0]
if machine_sn == current_sn:
machine.append(int(record.facility_id))
#print("First Machine in list: " + str(current_machines[0]))
evs = cyb_records.Events()
evs.ReadRecords(username=myusername, password=mypassword)
#clean_records = clean_events.CleanEvents(evs.records)
clean_records = evs.records
print("Total number of cleaned errors: " + str(len(clean_records)))
errors = []
# add in the number of errors by summing errors for each sn in the error table.
for record in clean_records:
machine_found = False
for machine in errors:
if machine[0] == str(record.sn):
machine[1] += 1
machine_found = True
if machine_found == False:
errors.append([str(record.sn), 1])
#print errors
# taking these numbers of errors and adding them to the current_machines list
for error in errors:
for machine in current_machines:
if machine[0] == error[0]:
machine.append(error[1])
# in case any machines have never thrown an error, make sure they still get an entry
for machine in current_machines:
if len(machine) < 4:
# machine has no prior errors
machine.append(0)
#print current_machines
# extract the variables into separate vectors for plotting, based on what we want to plot them by.
# mainly split up by site code- ignoring sites that don't exist.
real_up_time = []
real_error_count = []
fake_up_time = []
fake_error_count = []
for machine in current_machines:
site_code = machine[2]
if site_code >= 9 and site_code <= 12:
# site is a real one
real_up_time.append(machine[1])
real_error_count.append(machine[3])
else:
# site code is a fake. still worth trying?
fake_up_time.append(machine[1])
fake_error_count.append(machine[3])
print("Up times: " + str(real_up_time))
print("Error Counts: " + str(real_error_count))
# preparing to print with color codes for site codes. splitting up main vector again.
up_time9 = []
up_time10 = []
up_time11 = []
up_time12 = []
error_count9 = []
error_count10 = []
error_count11 = []
error_count12 = []
for machine in current_machines:
site_code = machine[2]
if site_code == 9:
up_time9.append(machine[1])
error_count9.append(machine[3])
elif site_code == 10:
up_time10.append(machine[1])
error_count10.append(machine[3])
elif site_code == 11:
up_time11.append(machine[1])
error_count11.append(machine[3])
elif site_code == 12:
up_time12.append(machine[1])
error_count12.append(machine[3])
else:
pass # fake machine
# plotting
#plt.scatter(real_up_time, real_error_count, s = 100)
#plt.scatter(fake_up_time, fake_error_count, s = 20)
# PUT THIS IN A FUNCTION!! NO NEED TO HAVE 10 copies!
up_time9_kilos = []
for time in up_time9:
up_time9_kilos.append(time / 1000.0)
up_time10_kilos = []
for time in up_time10:
up_time10_kilos.append(time / 1000.0)
up_time11_kilos = []
for time in up_time11:
up_time11_kilos.append(time / 1000.0)
up_time12_kilos = []
for time in up_time12:
up_time12_kilos.append(time / 1000.0)
#plt.scatter(up_time9_kilos, error_count9, facecolor= 'r', s = 300, label= 'site code 9')
#plt.scatter(up_time10_kilos, error_count10, facecolor= 'b', s = 300, label = 'site code 10')
#plt.scatter(up_time11_kilos, error_count11, facecolor= 'g', s = 300, label = 'site code 11')
#plt.scatter(up_time12_kilos, error_count12, facecolor= 'k', s = 300, label = 'site code 12')
# unfiltered version
plt.scatter(up_time9,
error_count9,
facecolor='r',
s=300,
label='site code 9')
plt.scatter(up_time10,
error_count10,
facecolor='b',
s=300,
label='site code 10')
plt.scatter(up_time11,
error_count11,
facecolor='g',
s=300,
label='site code 11')
plt.scatter(up_time12,
error_count12,
facecolor='k',
s=300,
label='site code 12')
plt.legend()
plt.xlabel('Distance (kilometers)', fontsize=24)
plt.ylabel('Number of Errors', fontsize=24)
plt.title('Errors vs. Distance, by Machine and Site', fontsize=30)
plt.xlim([0, 100000000])
plt.ylim([0, 600])
plt.show()