def plot_json(jsondir, measure_type, region, season=None, subset=None):
if season is None:
files = glob.glob('{0}*_{1}.json'.format(jsondir, measure_type))
else:
files = glob.glob('{0}*_{1}_{2}.json'.format(jsondir, measure_type,
season))
if measure_type == 'electric':
good_set = gbs.get_energy_set('eui_elec')
elif measure_type == 'gas':
good_set = gbs.get_energy_set('gas')
files = [f for f in files if len([x for x in good_set if x in f]) > 0]
print len(files)
if subset is not None:
files = [f for f in files if len([x for x in subset if x in f]) > 0]
# if season is None:
# files = glob.glob(os.getcwd() + \
# '/input/FY/interval/ion_0627/{0}/json_{1}/*_{2}.json'.format(dirname, occtime, measure_type))
# else:
# files = glob.glob(os.getcwd() + \
# '/input/FY/interval/ion_0627/{0}/json_{1}/*_{2}_{3}.json'.format(dirname, occtime, measure_type, season))
def get_name(string):
idx = string.rfind('/')
return string[idx + 1:idx + 9]
data = []
for x in files:
with open(x, 'r') as rd:
j = json.load(rd)
data.append(j)
data_str = 'series: [{0}]'.format(','.join(map(str, data)))
data_str = data_str.replace('u\'', '\'')
with open(
os.getcwd() +
'/input/FY/interval/ion_0627/piecewise_all/template.html',
'r') as rd:
lines = rd.readlines()
if season is None:
mytitle = 'Region {1} Monthly {0} (kBtu/sq.ft) vs Temperature (F)'.format(
measure_type, region)
else:
mytitle = '{1} Monthly {0} (kBtu/sq.ft) vs Temperature (F)'.format(
measure_type, season.title())
for i in range(len(lines)):
lines[i] = lines[i].replace('series: []', data_str)
lines[i] = lines[i].replace('Mytitle', mytitle)
lines[i] = lines[i].replace('Myylabel', 'kBtu/sq.ft')
if season is None:
f = os.getcwd(
) + '/plot_FY_weather/html/by_region/{}_piecewise_all_region_{}.html'.format(
measure_type, region)
if subset is None:
f = os.getcwd(
) + '/plot_FY_weather/html/by_region/{}_piecewise_all_region_{}_no_filter.html'.format(
measure_type, region)
with open(f, 'w+') as wt:
wt.write(''.join(lines))
print 'end'
return
def euas_covered():
covered = gbs.get_covered_set()
euas = gbs.get_all_building_set()
good_elec = gbs.get_energy_set('eui_elec')
good_gas = gbs.get_energy_set('eui_gas')
print len(covered.intersection(euas))
print len(covered.intersection(good_elec))
print len(covered.intersection(good_gas))
def euas_covered():
covered = gbs.get_covered_set()
euas = gbs.get_all_building_set()
good_elec = gbs.get_energy_set('eui_elec')
good_gas = gbs.get_energy_set('eui_gas')
print len(covered.intersection(euas))
print len(covered.intersection(good_elec))
print len(covered.intersection(good_gas))
def table_for_robust_set():
conn = uo.connect('all')
study_set = gbs.get_energy_set('eui').intersection(gbs.get_cat_set(['A', 'I'], conn))
df = pd.read_csv(os.getcwd() + '/plot_FY_weather/html/table/action_saving.csv')
df = df[df['Building_Number'].isin(study_set)]
df.sort('Building_Number', inplace=True)
df.to_csv(os.getcwd() + '/plot_FY_weather/html/table/action_saving_robustset.csv', index=False)
return
def study_set_plot():
conn = uo.connect('all')
with conn:
# df1 = pd.read_sql('SELECT DISTINCT Building_Number, Fiscal_Year FROM EUAS_monthly', conn)
df1 = pd.read_sql(
'SELECT DISTINCT Building_Number, Fiscal_Year FROM eui_by_fy',
conn)
df2 = pd.read_sql('SELECT Building_Number, Cat FROM EUAS_category',
conn)
meter_set = gbs.get_action_set('high_level_ECM', ['Advanced Metering'])
df = pd.merge(df1, df2, on='Building_Number', how='left')
df = df[df['Fiscal_Year'] > 2006]
df = df[df['Fiscal_Year'] < 2016]
df3 = df.groupby('Building_Number').filter(lambda x: len(x) > 5)
ai_set = gbs.get_cat_set(['A', 'I'], conn)
# invest = gbs.get_invest_set()[-1]
invest = gbs.get_ecm_set()
all_building = set(df3['Building_Number'].tolist())
print 'all building > 5 years of data: {0}'.format(len(all_building))
print 'all building > 5 years of data + ecm: {0}'.format(
len(all_building.intersection(invest)))
print 'all building > 5 years of data + meter: {0}'.format(
len(all_building.intersection(meter_set)))
df4 = df[df['Cat'].isin(
['A', 'I'])].groupby('Building_Number').filter(lambda x: len(x) > 5)
ai_building = set(df4['Building_Number'].tolist())
print 'A + I building > 5 years of data: {0}'.format(len(ai_building))
print 'A + I building > 5 years of data + ecm: {0}'.format(
len(ai_building.intersection(invest)))
print 'A + I building > 5 years of data + meter: {0}'.format(
len(ai_building.intersection(meter_set)))
print 'elec ', len(gbs.get_energy_set('elec').intersection(ai_set))
print 'elec + ecm', len(
gbs.get_energy_set('elec').intersection(ai_set).intersection(invest))
print 'elec + meter', len(
gbs.get_energy_set('elec').intersection(ai_set).intersection(
meter_set))
print 'gas ', len(gbs.get_energy_set('gas').intersection(ai_set))
print 'gas + ecm', len(
gbs.get_energy_set('gas').intersection(ai_set).intersection(invest))
print 'gas + meter', len(
gbs.get_energy_set('gas').intersection(ai_set).intersection(meter_set))
print 'eui', len(gbs.get_energy_set('eui').intersection(ai_set))
print 'eui + ecm', len(
gbs.get_energy_set('eui').intersection(ai_set).intersection(invest))
print 'eui + meter', len(
gbs.get_energy_set('eui').intersection(ai_set).intersection(meter_set))
return
def write_robust_energy_set():
s = gbs.get_energy_set('eui')
e = gbs.get_energy_set('eui_elec')
g = gbs.get_energy_set('eui_gas')
conn = uo.connect('all')
def get_status(x):
if x in s:
return "Electric EUI >= 12 and Gas EUI >= 3for at least 6 years from FY2007 to FY2015"
elif x in e:
return "Electric EUI >= 12 for at least 6 years from FY2007 to FY2015, there exist at least 4 years from FY2007 to FY2015 where Gas EUI < 3"
elif x in g:
return "Gas EUI >= 3 for at least 6 years from FY2007 to FY2015, there exist at least 4 years from FY2007 to FY2015 where Electric EUI < 12"
else:
return "There exist at least 4 years from FY2007 to FY2015 where Gas EUI < 3 and there exist at least 4 years from FY2007 to FY2015 where Electric EUI < 12"
with conn:
df = pd.read_sql('SELECT Building_Number, Cat FROM EUAS_category', conn)
# df['status'] = df['Building_Number'].map(lambda x: "Electric EUI >= 12 and Gas EUI >= 3for at least 6 years from FY2007 to FY2015" if x in s else 'Not robust energy')
df['status'] = df['Building_Number'].map(get_status)
df.to_csv(r_input + 'robust_energy_sep.csv', index=False)
def table_for_robust_set():
conn = uo.connect('all')
study_set = gbs.get_energy_set('eui').intersection(
gbs.get_cat_set(['A', 'I'], conn))
df = pd.read_csv(os.getcwd() +
'/plot_FY_weather/html/table/action_saving.csv')
df = df[df['Building_Number'].isin(study_set)]
df.sort('Building_Number', inplace=True)
df.to_csv(os.getcwd() +
'/plot_FY_weather/html/table/action_saving_robustset.csv',
index=False)
return
def write_robust_energy_set():
s = gbs.get_energy_set('eui')
e = gbs.get_energy_set('eui_elec')
g = gbs.get_energy_set('eui_gas')
conn = uo.connect('all')
def get_status(x):
if x in s:
return "Electric EUI >= 12 and Gas EUI >= 3for at least 6 years from FY2007 to FY2015"
elif x in e:
return "Electric EUI >= 12 for at least 6 years from FY2007 to FY2015, there exist at least 4 years from FY2007 to FY2015 where Gas EUI < 3"
elif x in g:
return "Gas EUI >= 3 for at least 6 years from FY2007 to FY2015, there exist at least 4 years from FY2007 to FY2015 where Electric EUI < 12"
else:
return "There exist at least 4 years from FY2007 to FY2015 where Gas EUI < 3 and there exist at least 4 years from FY2007 to FY2015 where Electric EUI < 12"
with conn:
df = pd.read_sql('SELECT Building_Number, Cat FROM EUAS_category',
conn)
# df['status'] = df['Building_Number'].map(lambda x: "Electric EUI >= 12 and Gas EUI >= 3for at least 6 years from FY2007 to FY2015" if x in s else 'Not robust energy')
df['status'] = df['Building_Number'].map(get_status)
df.to_csv(r_input + 'robust_energy_sep.csv', index=False)
def invest_cnt():
conn = uo.connect('all')
with conn:
df = pd.read_sql(
'SELECT DISTINCT Building_Number, high_level_ECM, detail_level_ECM FROM EUAS_ecm WHERE detail_level_ECM != \'GSALink\'',
conn)
eng_set = gbs.get_energy_set('eui')
ai_set = gbs.get_cat_set(['A', 'I'], conn)
study_set = eng_set.intersection(ai_set)
df = df[df['Building_Number'].isin(study_set)]
print df.groupby(['high_level_ECM', 'detail_level_ECM']).count()
print len(df)
df = df.groupby(['Building_Number']).filter(lambda x: len(x) == 1)
print len(df)
print df.head()
print df.groupby(['high_level_ECM', 'detail_level_ECM']).count()
return
def count_invest():
conn = uo.connect('all')
with conn:
df1 = pd.read_sql(
'SELECT DISTINCT Building_Number, high_level_ECM FROM EUAS_ecm WHERE high_level_ECM != \'GSALink\'',
conn)
df2 = pd.read_sql(
'SELECT DISTINCT Building_Number, ECM_program FROM EUAS_ecm_program',
conn)
eng_set = gbs.get_energy_set('eui')
ai_set = gbs.get_cat_set(['A', 'I'], conn)
study_set = eng_set.intersection(ai_set)
df1.dropna(subset=['high_level_ECM'], inplace=True)
df2.dropna(subset=['ECM_program'], inplace=True)
df1 = df1[df1['Building_Number'].isin(study_set)]
df2 = df2[df2['Building_Number'].isin(study_set)]
print df1['high_level_ECM'].value_counts()
print df2['ECM_program'].value_counts()
return
def fit_time(measure_type, region, season=None):
conn = uo.connect('all')
with conn:
df_bs = pd.read_sql(
'SELECT Building_Number, ICAO, eui_elec, eui_gas, year, month, ave FROM EUAS_monthly_weather',
conn)
df_region = pd.read_sql(
'SELECT DISTINCT Building_Number, [Region_No.] FROM EUAS_monthly',
conn)
if measure_type == 'electric':
good_set = gbs.get_energy_set('eui_elec')
elif measure_type == 'gas':
good_set = gbs.get_energy_set('gas')
df_bs = pd.merge(df_bs, df_region, on='Building_Number', how='left')
df_bs = df_bs[df_bs['Building_Number'].map(lambda x: x in good_set)]
df_bs = df_bs[df_bs['Region_No.'] == str(region)]
df_bs = df_bs[df_bs['eui_elec'].notnull()]
df_bs = df_bs[df_bs['eui_gas'].notnull()]
df_bs.sort_values(by=['Building_Number', 'year', 'month'],
ascending=False,
inplace=True)
bs_pair = list(set(zip(df_bs['Building_Number'], df_bs['ICAO'])))
df_bs['Timestamp'] = df_bs.apply(
lambda r: '{}-{}'.format(int(r['year']), int(r['month'])), axis=1)
sns.set_style("whitegrid")
sns.set_context("talk", font_scale=1)
ylabel = {'electric': 'electric (kBtu/sq.ft)', 'gas': 'gas kBtu/sq.ft'}
col_dict = {'electric': 'eui_elec', 'gas': 'eui_gas'}
print len(bs_pair)
sns.set_style("whitegrid")
# palette = sns.cubehelix_palette(len(bs_pair))
palette = sns.color_palette('husl', len(bs_pair))
sns.set_palette(palette)
colors_rgb = [util.float2hex(x) for x in palette]
sns.set_context("talk", font_scale=1)
jsondir = os.getcwd(
) + '/plot_FY_weather/html/by_region/Region{}/piecewise_all/json/'.format(
region)
title = "Region {}".format(region)
col = col_dict[measure_type]
for i, (b, s) in enumerate(bs_pair):
print b, s
df = df_bs[df_bs['Building_Number'] == b]
df = df.head(n=36)
print df.head()
points = df[col]
min_time = df['Timestamp'].min()
max_time = df['Timestamp'].max()
bx = plt.axes()
d0 = plot_piece(df,
bx,
title,
colors_rgb[i],
measure_type,
b,
s,
scatter=False,
annote=True,
jsondir=jsondir,
season=season)
plt.xlabel('Temperature_F')
# plt.show()
if season is None:
path = os.getcwd(
) + '/plot_FY_weather/html/by_region/Region{}/piecewise_all/{}.png'.format(
region, measure_type)
else:
path = os.getcwd(
) + '/plot_FY_weather/html/by_region/Region{}/piecewise_all/{}_{}.png'.format(
region, measure_type, season)
P.savefig(path,
dpi=my_dpi,
figsize=(2000 / my_dpi, 500 / my_dpi),
bbox_inches='tight')
plt.close()
return
