Skip to content

Commit

Permalink
Fixes bugs and miscomprehensions for rps
Browse files Browse the repository at this point in the history 
   eai by regions will come!
  • Loading branch information
spjuhel committed Jun 21, 2024
1 parent f7eaa01 commit e4c4dc9
Showing 1 changed file with 61 additions and 64 deletions.
125 changes: 61 additions & 64 deletions climada/engine/impact_trajectories.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -180,59 +180,62 @@ def calc_yearly_aais(yearly_eai_exp_0, yearly_eai_exp_1):
return yearly_aai_0, yearly_aai_1


def bayesian_mixer(start_snapshot, end_snapshot, metrics=["eai", "aai", "rp"], return_periods=[100, 500, 1000]):
# 1. Interpolate in between years
prop_H0, prop_H1 = bayesian_viktypliers(start_snapshot.year, end_snapshot.year)
imp_E0H0, imp_E1H0, imp_E0H1, imp_E1H1 = snapshot_combinaisons(
start_snapshot, end_snapshot
)
frequency_0 = start_snapshot.hazard.frequency
frequency_1 = end_snapshot.hazard.frequency

imp_mats_0 = interpolate_imp_mat(imp_E0H0, imp_E1H0, start_snapshot.year, end_snapshot.year)
imp_mats_1 = interpolate_imp_mat(imp_E0H1, imp_E1H1, start_snapshot.year, end_snapshot.year)

yearly_eai_exp_0, yearly_eai_exp_1 = calc_yearly_eais(imp_mats_0, imp_mats_1, frequency_0, frequency_1)

res = []

if "eai" in metrics:
yearly_eai = (
np.multiply(prop_H0.reshape(-1,1), yearly_eai_exp_0) +
np.multiply(prop_H1.reshape(-1,1), yearly_eai_exp_1)
)
eai_df = pd.DataFrame(index=list(range(start_snapshot.year, end_snapshot.year+1)),
data=yearly_eai)
eai_df["group"] = pd.NA
eai_df["metric"] = "eai"
eai_df.reset_index(inplace=True)
res.append(eai_df)

if "aai" in metrics:
yearly_aai_0, yearly_aai_1 = calc_yearly_aais(yearly_eai_exp_0, yearly_eai_exp_1)
yearly_aai = prop_H0 * yearly_aai_0 + prop_H1 * yearly_aai_1
aai_df = pd.DataFrame(index=list(range(start_snapshot.year, end_snapshot.year+1)),
data=yearly_aai)
aai_df["group"] = pd.NA
aai_df["metric"] = "aai"
aai_df.reset_index(inplace=True)
res.append(aai_df)

if "rp" in metrics:
tmp = []
for rp in return_periods:
rp_0, rp_1 = calc_yearly_rps(imp_mats_0, imp_mats_1, frequency_0, frequency_1, rp)
yearly_rp = np.multiply(prop_H0.reshape(-1,1), rp_0) + np.multiply(prop_H1.reshape(-1,1), rp_1)
tmp_df = pd.DataFrame(index=list(range(start_snapshot.year, end_snapshot.year+1)),
data=yearly_rp)
tmp_df["group"] = pd.NA
tmp_df["metric"] = f"rp_{rp}"
tmp_df.reset_index(inplace=True)
tmp.append(tmp_df)
rp_df = pd.concat(tmp)
res.append(rp_df)

return pd.concat(res)
def bayesian_mixer(start_snapshot, end_snapshot, metrics, return_periods):

Check warning on line 183 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

missing-function-docstring 

LOW:
Missing function or method docstring
Raw output 
no description found

Check warning on line 183 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

too-many-locals 

LOW:
Too many local variables (26/15)
Raw output 
Used when a function or method has too many local variables.
# 1. Interpolate in between years
prop_H0, prop_H1 = bayesian_viktypliers(start_snapshot.year, end_snapshot.year)

Check warning on line 185 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "prop_H0" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).

Check warning on line 185 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "prop_H1" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).
imp_E0H0, imp_E1H0, imp_E0H1, imp_E1H1 = snapshot_combinaisons(

Check warning on line 186 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "imp_E0H0" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).

Check warning on line 186 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "imp_E1H0" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).

Check warning on line 186 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "imp_E0H1" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).

Check warning on line 186 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

invalid-name 

LOW:
Variable name "imp_E1H1" doesn't conform to '(([a-z][a-z0-9_]{2,30})|(_[a-z0-9_]*))$' pattern
Raw output 
Used when the name doesn't match the regular expression associated to its type(constant, variable, class...).
start_snapshot, end_snapshot
)
frequency_0 = start_snapshot.hazard.frequency
frequency_1 = end_snapshot.hazard.frequency

imp_mats_0 = interpolate_imp_mat(imp_E0H0, imp_E1H0, start_snapshot.year, end_snapshot.year)
imp_mats_1 = interpolate_imp_mat(imp_E0H1, imp_E1H1, start_snapshot.year, end_snapshot.year)

yearly_eai_exp_0, yearly_eai_exp_1 = calc_yearly_eais(imp_mats_0, imp_mats_1, frequency_0, frequency_1)

Check warning on line 195 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

line-too-long 

LOW:
Line too long (107/100)
Raw output 
Used when a line is longer than a given number of characters.

res = []

year_idx = pd.Index(list (range(start_snapshot.year, end_snapshot.year+1)), name="year" )

# For the at_event needs rework
# if "eai" in metrics:
# yearly_eai = (
# np.multiply(prop_H0.reshape(-1,1), yearly_eai_exp_0) +
# np.multiply(prop_H1.reshape(-1,1), yearly_eai_exp_1)
# )
# eai_df = pd.DataFrame(index=year_idx,
# data=yearly_eai,
# columns=["result"]
# )
# eai_df["group"] = pd.NA
# eai_df["metric"] = "eai"
# eai_df.reset_index(inplace=True)
# res.append(eai_df)

if "aai" in metrics:
yearly_aai_0, yearly_aai_1 = calc_yearly_aais(yearly_eai_exp_0, yearly_eai_exp_1)
yearly_aai = prop_H0 * yearly_aai_0 + prop_H1 * yearly_aai_1
aai_df = pd.DataFrame(index=year_idx,
columns=["result"],
data=yearly_aai)
aai_df["group"] = pd.NA
aai_df["metric"] = "aai"
aai_df.reset_index(inplace=True)
res.append(aai_df)

if "rp" in metrics:
rp_0, rp_1 = calc_yearly_rps(imp_mats_0, imp_mats_1, frequency_0, frequency_1, return_periods)

Check warning on line 228 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

line-too-long 

LOW:
Line too long (102/100)
Raw output 
Used when a line is longer than a given number of characters.
yearly_rp = np.multiply(prop_H0.reshape(-1,1), rp_0) + np.multiply(prop_H1.reshape(-1,1), rp_1)

Check warning on line 229 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

line-too-long 

LOW:
Line too long (103/100)
Raw output 
Used when a line is longer than a given number of characters.
rp_df = pd.DataFrame(index=year_idx,
columns=return_periods,
data=yearly_rp).melt(value_name="result", var_name="rp", ignore_index=False)

Check warning on line 232 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

line-too-long 

LOW:
Line too long (110/100)
Raw output 
Used when a line is longer than a given number of characters.
rp_df.reset_index(inplace=True)
rp_df["group"] = pd.NA
rp_df["metric"] = f"rp_"+rp_df["rp"].astype(str)

Check warning on line 235 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

f-string-without-interpolation 

NORMAL:
Using an f-string that does not have any interpolated variables
Raw output 
no description found
res.append(rp_df)

return pd.concat(res,axis=0)

@dataclass
class Snapshot:

Check warning on line 241 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

missing-class-docstring 

LOW:
Missing class docstring
Raw output 
no description found
Expand Down Expand Up @@ -296,18 +299,12 @@ def calc_impacts_snapshots(self):
).impact()
return impacts_list

def calc_all_years(self):
all_yearly_aai = []
start = True
def calc_all_years(self, metrics=["eai", "aai", "rp"], return_periods=[100, 500, 1000]):

Check warning on line 302 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

dangerous-default-value 

NORMAL:
Dangerous default value [] as argument
Raw output 
Used when a mutable value as list or dictionary is detected in a default valuefor an argument.

Check warning on line 302 in climada/engine/impact_trajectories.py

View check run for this annotation

Jenkins - WCR / Pylint

missing-function-docstring 

LOW:
Missing function or method docstring
Raw output 
no description found
results_df = []
for start_snapshot, end_snapshot in pairwise(self.snapshots.data):
if start:
all_yearly_aai.append(self.bayesian_mixer(start_snapshot, end_snapshot))
start = False
else:
all_yearly_aai.append(
self.bayesian_mixer(start_snapshot, end_snapshot)[1:]
)
return np.concatenate(all_yearly_aai)
results_df.append(bayesian_mixer(start_snapshot, end_snapshot, metrics, return_periods))
results_df = pd.concat(results_df,axis=0)
return results_df[["group","year","metric","result"]]


#### WIP
Expand Down

0 comments on commit e4c4dc9

Please sign in to comment.