Merge pull request #55 from RSGInc/eschew-at-sign

at sign considered harmful

bca4abm/bca4abm.py

@@ -105,24 +105,23 @@ def assign_variables(assignment_expressions, df, locals_d):
     Evaluate a set of variable expressions from a spec in the context
     of a given data table.
 
-    There are two kinds of supported expressions: "simple" expressions are
-    evaluated in the context of the DataFrame using DataFrame.eval.
-    This is the default type of expression.
+    Expressions are evaluated using Python's eval function.
+    Python expressions have access to variables in locals_d (and df being
+    accessible as variable df.) They also have access to previously assigned
+    targets as the assigned target name.
 
-    Python expressions are evaluated in the context of this function using
-    Python's eval function. Because we use Python's eval this type of
-    expression supports more complex operations than a simple expression.
-    Python expressions are denoted by beginning with the @ character.
-    Users should take care that these expressions must result in
-    a Pandas Series.
+    Users should take care that expressions should result in
+    a Pandas Series (scalars will be automatically promoted to series.)
 
     Parameters
     ----------
-    exprs : sequence of str
+    assignment_expressions : pandas.DataFrame of target assignment expressions
+        target: target column names
+        expression: pandas or python expression to evaluate
     df : pandas.DataFrame
     locals_d : Dict
         This is a dictionary of local variables that will be the environment
-        for an evaluation of an expression that begins with @
+        for an evaluation of "python" expression.
 
     Returns
     -------
@@ -131,9 +130,10 @@ def assign_variables(assignment_expressions, df, locals_d):
 
     """
 
-    def to_series(x):
+    def to_series(x, target=None):
         if np.isscalar(x):
-            # assert False
+            if target:
+                print "WARNING: assign_variables promoting scalar %s to series" % target
             return pd.Series([x] * len(df), index=df.index)
         return x
 
@@ -148,9 +148,7 @@ def to_series(x):
         target = e[0]
         expression = e[1]
         try:
-            values = to_series(eval(expression[1:], globals(), locals_d)) \
-                if expression.startswith('@') \
-                else df.eval(expression)
+            values = to_series(eval(expression, globals(), locals_d), target=target)
             l.append((target, values))
 
             # FIXME - do we want to update locals to allows us to ref previously assigned targets?
@@ -159,10 +157,6 @@ def to_series(x):
             print "Variable %s expression failed for: %s" % (str(target), str(expression))
             raise err
 
-    # FIXME - should we add_assigned_columns rather than creating a dataframe and returning it?
-    # FIXME - we could pass in the target df to optionally assign directly from items
-    # FIXME - (though desired target might not be the eval df if eval df is a merged table...)
-
     # since we allow targets to be recycled, we want to only keep the last usage
     keepers = []
     for statement in reversed(l):

bca4abm/tests/configs/auto_ownership.csv

@@ -1,4 +1,4 @@
 Description,Target,Expression
-base scenario auto ownership cost,base_auto_ownership_cost, @persons.hh_expansion_factor * persons.base_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
-build scenario auto ownership cost,build_auto_ownership_cost, @persons.hh_expansion_factor * persons.build_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
-auto ownership cost benefit,diff_auto_ownership_cost, @build_auto_ownership_cost - base_auto_ownership_cost
+base scenario auto ownership cost,base_auto_ownership_cost, persons.hh_expansion_factor * persons.base_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
+build scenario auto ownership cost,build_auto_ownership_cost, persons.hh_expansion_factor * persons.build_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
+auto ownership cost benefit,diff_auto_ownership_cost, build_auto_ownership_cost - base_auto_ownership_cost

bca4abm/tests/configs/demographics.csv

@@ -1,5 +1,5 @@
 Target,Expression
 # compare hh_income to federal poverty threshold based on hh_size
-_poverty_threshold, @(persons.hh_size==1)*POVERTY_1 + (persons.hh_size>1)*POVERTY_2 + (persons.hh_size - 2).clip(lower=0)*POVERTY_N
-coc_poverty,@persons.hh_income <= _poverty_threshold
-coc_age, person_age > 65
+_poverty_threshold, (persons.hh_size==1)*POVERTY_1 + (persons.hh_size>1)*POVERTY_2 + (persons.hh_size - 2).clip(lower=0)*POVERTY_N
+coc_poverty, persons.hh_income <= _poverty_threshold
+coc_age, persons.person_age > 65

bca4abm/tests/configs/link.csv

@@ -1,20 +1,20 @@
 Description,Target,Expression
-,_auto_volume, total_volume - truck_volume
-,_vmt_auto, @ _auto_volume * links.distance
-,_vmt_truck, @ links.truck_volume * links.distance
+,_auto_volume, links.total_volume - links.truck_volume
+,_vmt_auto, _auto_volume * links.distance
+,_vmt_truck, links.truck_volume * links.distance
 # operating costs SANDAG version
-auto operating cost,cost_op_auto, @ _vmt_auto * OPERATING_COST_PER_MILE_AUTO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-truck operating cost,cost_op_truck, @ _vmt_truck * OPERATING_COST_PER_MILE_TRUCK * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-total auto and truck operating cost,cost_op_total, @ cost_op_auto + cost_op_truck
+auto operating cost,cost_op_auto, _vmt_auto * OPERATING_COST_PER_MILE_AUTO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+truck operating cost,cost_op_truck, _vmt_truck * OPERATING_COST_PER_MILE_TRUCK * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+total auto and truck operating cost,cost_op_total, cost_op_auto + cost_op_truck
 # unreliability costs SANDAG version
-,_time_ratio, congested_time / time
-,_ttim2, @ (TIME_RATIO_FACTOR * (_time_ratio ** TIME_RATIO_POWER)).clip(upper=TIME_RATIO_UPPER_LIMIT)
-,_tti50, @ _ttim2 ** TTI50_POWER
-,_tti80, @ 1 + log(_ttim2) * TTI80_LOG_MULTIPLIER
-,_dly_per_mile, @ _tti50 + (_tti80 - _tti50) * RELIABILITY_RATIO
-,_equiv_delay_auto, @ _dly_per_mile * links.distance * _auto_volume/60
-,_equiv_delay_truck, @ _dly_per_mile * links.distance * links.truck_volume/60
-auto delay cost,cost_delay_auto, @ _equiv_delay_auto * VALUE_OF_RELIABILITY_AUTO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-truck delay cost,cost_delay_truck, @ _equiv_delay_truck * VALUE_OF_RELIABILITY_TRUCK * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-total auto and truck cost,cost_delay_total, @ cost_delay_auto + cost_delay_truck
+,_time_ratio, links.congested_time / links.time
+,_ttim2, (TIME_RATIO_FACTOR * (_time_ratio ** TIME_RATIO_POWER)).clip(upper=TIME_RATIO_UPPER_LIMIT)
+,_tti50, _ttim2 ** TTI50_POWER
+,_tti80, 1 + log(_ttim2) * TTI80_LOG_MULTIPLIER
+,_dly_per_mile, _tti50 + (_tti80 - _tti50) * RELIABILITY_RATIO
+,_equiv_delay_auto, _dly_per_mile * links.distance * _auto_volume/60
+,_equiv_delay_truck, _dly_per_mile * links.distance * links.truck_volume/60
+auto delay cost,cost_delay_auto, _equiv_delay_auto * VALUE_OF_RELIABILITY_AUTO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+truck delay cost,cost_delay_truck, _equiv_delay_truck * VALUE_OF_RELIABILITY_TRUCK * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+total auto and truck cost,cost_delay_total, cost_delay_auto + cost_delay_truck
 

bca4abm/tests/configs/link_daily.csv

@@ -1,7 +1,7 @@
 Description,Target,Expression
-total vehicle miles travelled,vmt_total, total_volume * distance
+total vehicle miles travelled,vmt_total, links.total_volume * links.distance
 # accident costs SANDAG version based on daily volumes
-annualized crash cost property damage,crash_cost_pdo, @ vmt_total * CRASH_RATE_PDO * CRASH_COST_PDO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-annualized crash cost injury,crash_cost_injury, @ vmt_total * CRASH_RATE_INJURY * CRASH_COST_INJURY * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-annualized crash cost fatalities,crash_cost_fatal, @ vmt_total * CRASH_RATE_FATAL * CRASH_COST_FATAL * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-total annualized crash cost,crash_cost_total, @ crash_cost_pdo + crash_cost_injury + crash_cost_fatal
+annualized crash cost property damage,crash_cost_pdo, vmt_total * CRASH_RATE_PDO * CRASH_COST_PDO * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+annualized crash cost injury,crash_cost_injury, vmt_total * CRASH_RATE_INJURY * CRASH_COST_INJURY * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+annualized crash cost fatalities,crash_cost_fatal, vmt_total * CRASH_RATE_FATAL * CRASH_COST_FATAL * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+total annualized crash cost,crash_cost_total, crash_cost_pdo + crash_cost_injury + crash_cost_fatal

bca4abm/tests/configs/person_trips.csv

@@ -1,17 +1,17 @@
 Description,Target,Expression
-,_work, @trips.tour_purpose.map(TOUR_PURPOSE_WORK_MAP)
+,_work, trips.tour_purpose.map(TOUR_PURPOSE_WORK_MAP)
 #
-monetized auto ivt,auto_time, @-0.5 * trips.hh_expansion_factor * (trips.build_auto_time - trips.base_auto_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-monetized transit ivt,transit_time, @-0.5 * trips.hh_expansion_factor * (trips.build_transit_time - trips.base_transit_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-monetized transit wait time,transit_wait_time, @-0.5 * trips.hh_expansion_factor * (trips.build_transit_wait - trips.base_transit_wait) * _work.map(WORK_WAIT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-monetized transit walk time,transit_walk_time, @-0.5 * trips.hh_expansion_factor * (trips.build_transit_walk - trips.base_transit_walk) * _work.map(WORK_WALK_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-monetized bike time,bike_time, @-0.5 * trips.hh_expansion_factor * (trips.build_bike_time - trips.base_bike_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-monetized walk time,walk_time, @-0.5 * trips.hh_expansion_factor * (trips.build_walk_time - trips.base_walk_time) * _work.map(WORK_WALK_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-toll cost,toll, @-0.5 * trips.hh_expansion_factor * (trips.build_toll_cost - trips.base_toll_cost) * TOLL_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-fuel cost,fuel, @-0.5 * trips.hh_expansion_factor * (trips.build_fuel_cost - trips.base_fuel_cost) * FUEL_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-park cost,park, @-0.5 * trips.hh_expansion_factor * (trips.build_park_cost - trips.base_park_cost) * PARK_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
-fare cost,fare, @-0.5 * trips.hh_expansion_factor * (trips.build_fare_cost - trips.base_fare_cost) * FARE_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized auto ivt,auto_time, -0.5 * trips.hh_expansion_factor * (trips.build_auto_time - trips.base_auto_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized transit ivt,transit_time, -0.5 * trips.hh_expansion_factor * (trips.build_transit_time - trips.base_transit_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized transit wait time,transit_wait_time, -0.5 * trips.hh_expansion_factor * (trips.build_transit_wait - trips.base_transit_wait) * _work.map(WORK_WAIT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized transit walk time,transit_walk_time, -0.5 * trips.hh_expansion_factor * (trips.build_transit_walk - trips.base_transit_walk) * _work.map(WORK_WALK_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized bike time,bike_time, -0.5 * trips.hh_expansion_factor * (trips.build_bike_time - trips.base_bike_time) * _work.map(WORK_IVT_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+monetized walk time,walk_time, -0.5 * trips.hh_expansion_factor * (trips.build_walk_time - trips.base_walk_time) * _work.map(WORK_WALK_VOT_MAP)/60.0 * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+toll cost,toll, -0.5 * trips.hh_expansion_factor * (trips.build_toll_cost - trips.base_toll_cost) * TOLL_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+fuel cost,fuel, -0.5 * trips.hh_expansion_factor * (trips.build_fuel_cost - trips.base_fuel_cost) * FUEL_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+park cost,park, -0.5 * trips.hh_expansion_factor * (trips.build_park_cost - trips.base_park_cost) * PARK_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
+fare cost,fare, -0.5 * trips.hh_expansion_factor * (trips.build_fare_cost - trips.base_fare_cost) * FARE_INCLUSION_FACTOR * DISCOUNT_RATE * ANNUALIZATION_FACTOR
 #
-,_monetized_time, @auto_time+transit_time+transit_wait_time+transit_walk_time+bike_time+walk_time
-,_cost, @toll+fuel+park+fare
-total benefit,total, @(_monetized_time + _cost)
+,_monetized_time, auto_time+transit_time+transit_wait_time+transit_walk_time+bike_time+walk_time
+,_cost, toll+fuel+park+fare
+total benefit,total, _monetized_time + _cost

bca4abm/tests/configs/physical_activity_person.csv

@@ -1,15 +1,15 @@
 Description,Target,Expression
-,_under_20, person_age < 20
-,_over_65, person_age >= 65
-,_over_75, person_age >= 75
-,_walk_risk_factor, @WALK_FACTOR_AGE_UNDER_20 * _under_20 + WALK_FACTOR_AGE_20_TO_74 * ( ~ (_under_20 | _over_75)) + WALK_FACTOR_AGE_75_PLUS * _over_75
-,_bike_risk_factor, @BIKE_FACTOR_AGE_UNDER_20 * _under_20 + BIKE_FACTOR_AGE_20_TO_64 * ( ~ (_under_20 | _over_65)) + BIKE_FACTOR_AGE_65_PLUS * _over_65
-,_base_walk_risk_reduction, @(_walk_risk_factor * persons.base_walk / WALK_DIVISOR_MINUTES_PER_DAY).clip(upper=WALK_MAX_RISK_REDUCTION)
-,_base_bike_risk_reduction, @(_bike_risk_factor * persons.base_bike / BIKE_DIVISOR_MINUTES_PER_DAY).clip(upper=BIKE_MAX_RISK_REDUCTION)
-,_base_total_risk_reduction, @(_base_walk_risk_reduction+_base_bike_risk_reduction).clip(upper=TOTAL_MAX_RISK_REDUCTION)
-base scenario value of risk reduction,base_value_of_risk_reduction, @_base_total_risk_reduction * VALUE_OF_MORTALITY_RISK_PER_YEAR * persons.hh_expansion_factor
-,_build_walk_risk_reduction, @(_walk_risk_factor * persons.build_walk / WALK_DIVISOR_MINUTES_PER_DAY).clip(upper=WALK_MAX_RISK_REDUCTION)
-,_build_bike_risk_reduction, @(_bike_risk_factor * persons.build_bike / BIKE_DIVISOR_MINUTES_PER_DAY).clip(upper=BIKE_MAX_RISK_REDUCTION)
-,_build_total_risk_reduction, @(_build_walk_risk_reduction+_build_bike_risk_reduction).clip(upper=TOTAL_MAX_RISK_REDUCTION)
-build scenario value of risk reduction,build_value_of_risk_reduction, @_build_total_risk_reduction * VALUE_OF_MORTALITY_RISK_PER_YEAR * persons.hh_expansion_factor
-benefit value of risk reduction,benefit_risk_reduction, @build_value_of_risk_reduction - base_value_of_risk_reduction
+,_under_20, persons.person_age < 20
+,_over_65, persons.person_age >= 65
+,_over_75, persons.person_age >= 75
+,_walk_risk_factor, WALK_FACTOR_AGE_UNDER_20 * _under_20 + WALK_FACTOR_AGE_20_TO_74 * ( ~ (_under_20 | _over_75)) + WALK_FACTOR_AGE_75_PLUS * _over_75
+,_bike_risk_factor, BIKE_FACTOR_AGE_UNDER_20 * _under_20 + BIKE_FACTOR_AGE_20_TO_64 * ( ~ (_under_20 | _over_65)) + BIKE_FACTOR_AGE_65_PLUS * _over_65
+,_base_walk_risk_reduction, (_walk_risk_factor * persons.base_walk / WALK_DIVISOR_MINUTES_PER_DAY).clip(upper=WALK_MAX_RISK_REDUCTION)
+,_base_bike_risk_reduction, (_bike_risk_factor * persons.base_bike / BIKE_DIVISOR_MINUTES_PER_DAY).clip(upper=BIKE_MAX_RISK_REDUCTION)
+,_base_total_risk_reduction, (_base_walk_risk_reduction+_base_bike_risk_reduction).clip(upper=TOTAL_MAX_RISK_REDUCTION)
+base scenario value of risk reduction,base_value_of_risk_reduction, _base_total_risk_reduction * VALUE_OF_MORTALITY_RISK_PER_YEAR * persons.hh_expansion_factor
+,_build_walk_risk_reduction, (_walk_risk_factor * persons.build_walk / WALK_DIVISOR_MINUTES_PER_DAY).clip(upper=WALK_MAX_RISK_REDUCTION)
+,_build_bike_risk_reduction, (_bike_risk_factor * persons.build_bike / BIKE_DIVISOR_MINUTES_PER_DAY).clip(upper=BIKE_MAX_RISK_REDUCTION)
+,_build_total_risk_reduction, (_build_walk_risk_reduction+_build_bike_risk_reduction).clip(upper=TOTAL_MAX_RISK_REDUCTION)
+build scenario value of risk reduction,build_value_of_risk_reduction, _build_total_risk_reduction * VALUE_OF_MORTALITY_RISK_PER_YEAR * persons.hh_expansion_factor
+benefit value of risk reduction,benefit_risk_reduction, build_value_of_risk_reduction - base_value_of_risk_reduction

bca4abm/tests/configs/physical_activity_trip.csv

@@ -1,5 +1,5 @@
 Target,Expression
-base_walk, (base == 1) * (base_transit_walk + base_walk_time)
-base_bike, (base == 1) * base_bike_time
-build_walk, (build == 1) * (build_transit_walk + build_walk_time)
-build_bike, (build == 1) * build_bike_time
+base_walk, (trips.base == 1) * (trips.base_transit_walk + trips.base_walk_time)
+base_bike, (trips.base == 1) * trips.base_bike_time
+build_walk, (trips.build == 1) * (trips.build_transit_walk + trips.build_walk_time)
+build_bike, (trips.build == 1) * trips.build_bike_time

example/configs/auto_ownership.csv

@@ -1,4 +1,4 @@
 Description,Target,Expression
-base scenario auto ownership cost,base_auto_ownership_cost, @persons.hh_expansion_factor * persons.base_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
-build scenario auto ownership cost,build_auto_ownership_cost, @persons.hh_expansion_factor * persons.build_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
-auto ownership cost benefit,diff_auto_ownership_cost, @build_auto_ownership_cost - base_auto_ownership_cost
+base scenario auto ownership cost,base_auto_ownership_cost, persons.hh_expansion_factor * persons.base_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
+build scenario auto ownership cost,build_auto_ownership_cost, persons.hh_expansion_factor * persons.build_vehicles * ANNUAL_COST_PER_VEHICLE / persons.hh_size
+auto ownership cost benefit,diff_auto_ownership_cost, build_auto_ownership_cost - base_auto_ownership_cost

example/configs/demographics.csv

@@ -1,5 +1,5 @@
 Target,Expression
 # compare hh_income to federal poverty threshold based on hh_size
-_poverty_threshold, @(persons.hh_size==1)*POVERTY_1 + (persons.hh_size>1)*POVERTY_2 + (persons.hh_size - 2).clip(lower=0)*POVERTY_N
-coc_poverty,@persons.hh_income <= _poverty_threshold
-coc_age, person_age > 65
+_poverty_threshold, (persons.hh_size==1)*POVERTY_1 + (persons.hh_size>1)*POVERTY_2 + (persons.hh_size - 2).clip(lower=0)*POVERTY_N
+coc_poverty, persons.hh_income <= _poverty_threshold
+coc_age, persons.person_age > 65