The Data Space#

Input data shared by every by-contract projection.

The seven input CSVs are read here, once per model, and referenced from Projection as data. Projection is parameterized by idx, the position of a model point in the table, so each Projection[i] is a separate ItemSpace with its own cells cache; if the readers lived there, every model point would re-read every file. Holding them in an unparameterized Space reads each file once no matter how many contracts are projected.

Inputs are external files: plain CSVs in the model folder’s parent directory, products/fixed_deferred_annuity/, rather than data stored inside the model. The model folder therefore holds nothing but formulas — no _data/, no IOSpec, no embedded values — so a diff of the model shows logic changes only. This follows annuallife.TradLife_A; contrast basiclife.BasicTerm_S, which keeps its inputs inside the model through modelx’s IOSpec machinery.

The consequence worth knowing: the model is not portable on its own. Copying the MYGA_US_SEC folder without its parent’s CSVs produces a model that reads and then fails on first evaluation.

Six of the tables are MYGA_US_S’s. The mortality table, the surrender charge schedules and their attained-age cap, the rate scenarios, the withdrawal programmes and the declared-differential MVA factors are the files MYGA_US_S reads, read here in place: one copy of each table serves both models, so an edit to one reaches both. The model point table is this model’s own, model_point_table_sec.csv; it sits beside MYGA_US_S’s model_point_table.csv, which this model never opens.

input_dir() resolves the directory from _model.path.parent at run time, so the model works wherever the repository is checked out. Each table has a filename Reference and a reader Cells:

Reference

Cells

File

model_point_file

model_point_table()

model_point_table_sec.csv

mort_table_file

mort_table()

mort_table.csv

surr_charge_file

surr_charge_table()

surr_charge_table.csv

surr_charge_age_cap_file

surr_charge_age_cap_table()

surr_charge_age_cap.csv

rate_scenario_file

rate_scenario()

rate_scenario.csv

withdrawal_file

withdrawal_table()

withdrawal_table.csv

mva_factor_file

mva_factor_table()

mva_factor_table.csv

The model point table is checked as it is read. Twelve of its columns are the in-force interface: policy_count and duration_mth, as in basiclife.BasicTerm_SE, and the ten columns of an in-force contract’s opening state. model_point_table() refuses a table that lacks any of them, repeats a point_id, holds a count or a duration no contract can have, holds an opening state no contract can be in, or holds a key that the table it joins lacks: a scenario that sets no level by the point’s first month on the books, a withdrawal programme, or a sex and an attained age the mortality table gives no rate at. The checks run in turn, and the first that fails raises, naming the missing column or every point that fails it. It also refuses a contract to be issued more than future_horizon_mth months after the valuation date, 120 months as shipped: future new business is written over a stated horizon, so that no frame runs longer than cover from issue plus that horizon. Set to 0, the Reference refuses future new business altogether. Whether a point is still in cover at the valuation date is a rule of the product rather than of the table, and Projection.proj_len checks it.

So are the keys of the six assumption tables. Each table’s reader above is MYGA_US_S’s, unchanged, and beside it a cell reads the table’s keys by value, once, checks them and its values, and returns the table indexed by the checked keys: mort_rates(), surr_charge_scales(), surr_charge_age_caps(), rate_scenario_steps(), wd_schedules() and mva_factors(). The projection reads each table there and nowhere else, and takes the age cap’s extents, surr_charge_age_cap_first() and surr_charge_age_cap_last(), and the duration factors’ last year, mva_factor_last_year(), from the same checked keys, so a projection meets each table’s check before it reads a value. A key that is not a whole number, text and a blank among them, is refused, naming the file, the column and the lines it is on, and so are a blank text key, a value that is not a number in its range, a repeated key, and, where the projection reads every year or age between a table’s first and last, keys that do not run one row a year or an age, and so is a surrender charge table without both the schedules the projection reads. A key written as a float, 1.0, is the whole number it holds, as MYGA_US_S’s lookups match it. Those lookups take the keys as pandas reads them: one key written as text turns its whole column to text, so that every lookup misses, and a surrender charge or a withdrawal is then zero without a word, or the read raises a KeyError that names no file.

Four of the tables carry a compound key. surr_charge_table is indexed by (schedule, contract_year) so the initial 9/8/7/6/5 schedule [S10] and the renewal 5/4/3/2/1 schedule [S2] sit in one file; mort_table by (age, sex); rate_scenario by (scenario_id, t), read as a step function of t, and withdrawal_table by (wd_schedule_id, t), an exact-key lookup, so a scenario or a withdrawal programme is a handful of rows rather than one row per month. The two t are on different clocks. A scenario’s t is a valuation month: the scenario is a path in calendar time, shared by every contract of a block, and a row keyed t = 12 holds from twelve months after the valuation date. A withdrawal’s t is a policy month, read at duration_mth(t): a row keyed t = 12 falls in the first month of contract year 2, however old the contract is at the valuation date.

To swap in a licensed mortality basis — the 2012 IAM Basic table with Projection Scale G2 and the VM-22 Table 6.7 factors that the notes prescribe and that may not be redistributed here — replace mort_table.csv with a same-schema file, or point mort_table_file at a different name, before the model computes anything: a compiled model clears no cache, so there the name is set on a package just imported, before anything is read. No formula changes. Replacing a file changes MYGA_US_S too, which reads the same ones.

The vectors. The readers and the checked tables return what the files hold, and no Projection cell but model_point, a reporting cell, reads them. The others read the arrays below instead, one element at a time, a model point’s attribute being the element at its position, data.premium()[idx]. That is what lets the model compile to C, as basiclife.BasicTerm_SC does: each array is a numpy vector or table of int64 or float64, which compiles to a typed memoryview, never one of bools, objects or text, and each count and extent is an int.

Text the projection needs is decoded here, into the number it selects: each value of the five categorical columns MYGA_US_S branches on into a 0/1 flag, the is_* cells, and free_wd_mva_exempt into another. Each table keyed by more than a number becomes an integer-indexed array: the scenarios, the withdrawal programmes and the mortality table are numbered, each point carrying the number of its own, scenario_number, wd_schedule_number and mort_schedule, all three numbered from 1 in the sorted order of their keys; the scenarios are step-filled by valuation month, the programmes held by policy month, the surrender charges by schedule code and contract year, the age cap and the mortality table by attained age, padded from 0, and the duration factors by whole year remaining, so that the month, the year or the age is the index:

Cells

dtype

Shape

From

point_id()

int64

[point]

the table’s index

age_at_entry()

int64

[point]

age_at_entry

duration_mth()

int64

[point]

duration_mth

guar_period()

int64

[point]

guar_period

policy_count()

float64

[point]

policy_count

premium() .. premium_tax_rate()

float64

[point]

eight columns, read by float()

av_pp_init() and nine more

float64

[point]

the ten columns of the opening state, as model_point_table() lists them

free_wd_mva_exempt()

int64

[point]

bool() of the value

is_wd_gross() .. is_cap_gmir_floor()

int64

[point]

the fourteen decodes, 1 on the value

scenario_count()

int

scalar

rate_scenario.csv

scenario_number()

int64

[point]

scenario_id, joined

scenario_first_mths()

int64

[scenario]

each first t, 0 or more

scenario_last_mth()

int

scalar

the last t, 0 or more

mva_ref_yield_array()

float64

[scenario, t]

mva_ref_yield, step-filled

market_rate_array()

float64

[scenario, t]

market_rate, step-filled

wd_schedule_count()

int

scalar

withdrawal_table.csv

wd_schedule_number()

int64

[point]

wd_schedule_id, joined

wd_last_mth()

int

scalar

the last policy month, 0 or more

wd_amount_array()

float64

[schedule, month]

wd_amount, 0.0 without a row

surr_charge_last_years()

int64

[3]

each schedule’s last year

surr_charge_array()

float64

[3, year]

surr_charge_table.csv

surr_charge_age_cap_array()

float64

[age]

the age cap’s file

mort_schedule_count()

int

scalar

mort_table.csv

mort_schedule()

int64

[point]

sex, joined

mort_age_firsts(), mort_age_lasts()

int64

[schedule]

each sex’s ages

mort_table_array()

float64

[schedule, age]

mort_table.csv

mva_factor_array()

float64

[year, column]

mva_factor_table.csv

The table checks are what they were, model_point_table() and the six checked tables, beside the three int extents, surr_charge_age_cap_first(), surr_charge_age_cap_last() and mva_factor_last_year(), and one more, model_point_numbers(), which reads the eighteen numeric columns as MYGA_US_S reads them. A scenario array or a programme’s is as wide as its table’s last month, scenario_last_mth() or wd_last_mth(), and no wider than read_last_mth, the last month any read reaches: 1,321 as shipped, the last element of Projection’s t cache, 12 x (maturity_age - issue_age_min) + future_horizon_mth + 1, which spec_sec.py derives, since a scenario is read at a valuation month in the frame or the month after it, and a programme at a policy month in cover, 1,199 at the most. A row past it is held nowhere: no read reaches it, and the step function holds the last level at or before the read up to it, so a key far past any frame, which MYGA_US_S looks up and never reaches, neither widens an array nor refuses a point MYGA_US_S projects. Data cannot read Projection’s References, so a change to maturity_age, issue_age_min or future_horizon_mth needs read_last_mth recomputed with the spec, and a test holds the two to each other. The surrender charge rows are numbered as Projection’s schedule codes are, initial 1 and renewal 2, since Data cannot read Projection’s References; a test holds the two to each other.

An array is read by element, never whole, in this Space as in Projection. The compiler types an array cell as a memoryview only once it has seen that every use of it, in either Space, is an element read: one len(), slice, loop over it or numpy call on it anywhere, and the cell compiles as a Python object. So an array derived here from another reads the table’s column, or the other cell element by element, and a width comes from a table or a count cell, never from another cell’s array. Each table’s check is one cell that every array, extent and number of the table reads first, so that a projection, which reads an extent or a number first, meets the same refusal as a reader of the array. A keyword named as a cell of this Space would be rewritten by the export, so none is passed, and a local of a formula is named apart from the cells.

What cannot be decoded is refused, when the vector is first read, naming every point at fault, where it is not MYGA_US_S’s reading: a point_id that is not a whole number, and text in one of the eighteen numeric columns (model_point_numbers()), where MYGA_US_S raises at the first month that reads it. A blank in one of them is NaN and projects NaN, as there. A decode compares the text exactly, element by element, as MYGA_US_S compares it, so a value written in another case or with a space, Pct_av, or pct_av with a leading space, is no flag, and the projection answers as MYGA_US_S answers it, raising invalid free_wd_rule where MYGA_US_S raises it, or taking its else branch. free_wd_mva_exempt is MYGA_US_S’s bool() of the value, so a blank, read as NaN, and any text, False written in a file pandas reads as text among them, are True: kept, as the reading MYGA_US_S projects. A join number maps the point’s key as MYGA_US_S’s .loc matches it, by the value pandas read, and model_point_table() has refused a key its table lacks. A read beyond a table’s extent is refused where MYGA_US_S’s lookup raised, by the one Projection cell that reads the table, naming the point: an attained age outside the sex’s in mort_rate_at and a month before the scenario’s first in scenario_rate. A year past a surrender charge schedule’s last, and a month a programme has no row at, are no charge and no withdrawal, as there.

An int64 holds what it is read from. A float past int64 casts to the least int64 without a word, so a number read into an int64 vector is first refused unless it is below 2**53 in size, where the int64 holds MYGA_US_S’s reading of it: the issue age, duration_mth and guar_period (model_point_table()), the point_id (point_id()), and each assumption table’s key, refused by the table’s check, naming the file, the column and the lines: an age, a contract year, a month or a year remaining. Each is a whole number already, so no fraction truncates. A key past int64 would otherwise be read as the least int64, a scenario’s first level from month 0 among them, where MYGA_US_S refuses the point. An age table is held from age 0, so a negative age is refused, in mort_rates() and surr_charge_age_caps(), though MYGA_US_S projects a point on such a table, never reading the row.

Cells Descriptions#

input_dir()[source]#

The directory holding the input CSVs: the model folder’s parent.

Inputs are external files, not data stored inside the model, so the model folder is pure formulas. The path is resolved at run time from where the model was read, following annuallife.TradLife_A.

model_point_table()[source]#

The model point table, read from model_point_table_sec.csv and checked.

Indexed by point_id. Twelve columns are the in-force interface: policy_count and duration_mth, the completed policy months at the valuation date, as in basiclife.BasicTerm_SE, and the ten columns of the opening state, per contract, as the contract stands at the start of month 0, before any month-0 event:

  • av_pp_init, mgsv_pp_init and prem_accum_gmir_pp_init: the account value, the Model #805 floor and P_accum@GMIR entering month 0;

  • interest_credited_pp_init: the interest credited from issue to the valuation date;

  • interest_credited_pp_prev_anniv_init: the same to the anniversary on or before the valuation date, the one that opened the contract year under way;

  • free_wd_allow_init and free_wd_remain_init: that year’s free-withdrawal allowance, and the part of it unused entering month 0;

  • tax_basis_pp_init: the IRC 72 basis entering month 0, or at issue for a contract not yet in force; MYGA_US_S’s tax_basis_initial, renamed;

  • credit_rate_init and mva_ref_yield_locked_init: the declared rate and the MVA reference yield i0 in force entering month 0.

The rest are MYGA_US_S’s, except row_kind, source_policy_id and source_mth, which record where the generator took each shipped row from and which no formula reads.

The checks run over the whole table at once, and the first that fails raises ValueError, naming the missing column or every point it fails on:

  • the twelve in-force columns are present;

  • no point_id appears twice;

  • policy_count is a number, zero or more;

  • duration_mth is a whole number of months: positive for a contract in force, 0 for one issued at the valuation date, negative for one issued later;

  • age_at_entry and guar_period are whole numbers of years. MYGA_US_S reads both by int(), which would truncate a fraction without a word, and the initial-term check below reads guar_period as written;

  • duration_mth, age_at_entry and guar_period are each below 2**53 in size, where the int64 vectors that hold them, duration_mth(), age_at_entry() and guar_period(), hold each exactly (the Space docstring, An int64 holds what it is read from);

  • guar_period is 1 year or more: at 0 the guarantee-period index divides by zero, naming no point, and below it the projection runs a period no contract has;

  • no contract is issued more than future_horizon_mth months after the valuation date, so duration_mth is -future_horizon_mth or more. At 0 the model takes no future new business, and the message says so; a negative horizon is itself refused;

  • the account value, the two interest columns, the allowance, the remainder and the basis are each a finite number, zero or more;

  • the floor, P_accum, the declared rate and i0 are each a finite number. The first two may be negative: every withdrawal is deducted from both, and nothing floors either;

  • a contract not yet in force, duration_mth 0 or less, has no opening state: every state column is 0 but tax_basis_pp_init, its basis at issue;

  • a contract in force, duration_mth 1 or more, has an account value above 0: one fully withdrawn is no longer in force, and the dynamic lapse, which reads the surrender benefit as a share of the account value, would have nothing to divide by;

  • in the first contract year, duration_mth 1 to 11, the year opened at issue with nothing credited, so interest_credited_pp_prev_anniv_init is 0, and free_wd_allow_init is the issue allowance, free_wd_rate x premium under pct_av and 0 under interest_only;

  • on a point valued on a policy anniversary, duration_mth a multiple of 12, the interest at the anniversary equals the interest entering month 0: the two are one instant;

  • the interest at the anniversary does not exceed the interest to the valuation date;

  • under interest_only, from the second contract year on, duration_mth 12 or more, the year’s allowance, the interest credited over the year before, does not exceed the interest credited to the anniversary that opened the year, and in the second contract year, duration_mth 12 to 23, it is that interest, the first year’s;

  • off an anniversary, the remainder does not exceed the year’s allowance;

  • entering a month of the initial guarantee period, duration_mth 1 to 12 x guar_period, the first renewal month included, the declared rate and i0 are declared_rate_initial and mva_ref_yield_at_issue: the rate entering that month is still the initial one;

  • after it, the declared rate is the GMIR or more;

  • on a point valued on a policy anniversary under pct_av, the year’s allowance is free_wd_rate times the account value entering month 0, since the anniversary fixes the year’s base at that value;

  • last, each point’s keys into the tables it is joined with: its scenario_id is a scenario of rate_scenario_steps() that sets its levels by the first month the point is on the books, t = 0 or its issue month if later; its wd_schedule_id is a programme of wd_schedules(); and mort_rates() gives a rate at its sex and its attained age in that first month. So reading the table reads those three tables, checked.

Amounts of money are compared to a cent, so that an extract may carry each to the cent, as an administration system holds it. Each amount so carried is within half a cent of its exact value, so a rule that holds between the exact amounts can miss by half a cent from each: from the allowance, and from the base it is a share of, the account value or the premium, scaled by free_wd_rate, 0.0055 in all at the product’s 10%. A cent holds every such pair with room for the floating-point error of the comparison, where half a cent left the ties to it: point 11’s allowance and account value, 12,432.034927 and 124,320.349270, carried as 12,432.03 and 124,320.35, are 0.005 apart, and so is an issue allowance carried as 864.18 or 864.19 at 7% of a premium of 12,345.50, which is 864.185. So are read an allowance of 10,445.00 on an account value of 104,449.99, and one of 7,000.00 at 7% of 100,000, which is 7,000.000000000001 in floating point. A stale allowance, last year’s or one a dollar out, is still refused.

Accepted, because a contract can be in each state: a negative floor or P_accum; any basis at issue on a contract not yet in force, 0 for a qualified one; on a point valued on an anniversary a remainder above the year’s allowance, last year’s, which lapses there unread; and a duration_mth, an age_at_entry or a guar_period written as a float, read as the whole number it holds.

Left to the projection: an attained age past the mortality table’s last before the deemed maturity, because the maturity age is a Projection Reference, which Data cannot read. Projection.proj_len refuses such a point against mort_age_lasts(), where MYGA_US_S raises a KeyError once the projection reaches the age. The shipped table runs to 120, past every contract’s cover. So is the issue age the contract admits, 0 to 85, Projection.issue_age_min and issue_age_max. And text in a column the projection reads as a number is refused by model_point_numbers().

mort_table()[source]#

Annual mortality by attained age and sex, read from mort_table.csv.

surr_charge_table()[source]#

Surrender charge rates by schedule and contract year.

Read from surr_charge_table.csv.

surr_charge_age_cap_table()[source]#

The attained-age cap on the renewal surrender charge.

Read from surr_charge_age_cap.csv, indexed by attained age.

rate_scenario()[source]#

The exogenous rate scenarios, read from rate_scenario.csv.

Indexed by (scenario_id, t), t a valuation month, 0 at the valuation date, and read as a step function of t: each row states the MVA reference yield and the market competitor rate that hold from that month until the next row of the same scenario, so a flat path is one row at t = 0. MYGA_US_S reads the same rows on its own t, which on a contract issued at the valuation date is the same month.

withdrawal_table()[source]#

Scheduled gross withdrawals, read from withdrawal_table.csv.

Indexed by (wd_schedule_id, t), t a policy month, which the projection reads at duration_mth(t); a month with no row takes no withdrawal. The worked example’s $4,000 sits at t = 12, the first month of contract year 2, however old the contract is at the valuation date.

mva_factor_table()[source]#

The declared-differential MVA duration factors, from mva_factor_table.csv.

mort_rates()[source]#

Annual mortality by attained age and sex, as mort_table() reads it, checked.

Indexed by (age, sex), the age read by value, once, so that an age written 61.0 is 61, as MYGA_US_S’s lookup matches it. The arrays and extents of Projection.mort_rate_at, the table’s one reader, read it here, so a projection meets the check first. An age that is not a whole number, text and a blank among them, a negative age, a blank sex, a rate that is not a number from 0 to 1, and a sex whose ages do not run one row an age, without a gap or a repeat, are refused, naming the file, the column and the lines they are on. A negative age is refused because mort_table_array() holds the table from age 0, where MYGA_US_S projects a point on such a table, never reading the row.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

surr_charge_scales()[source]#

The surrender charge schedules, as surr_charge_table() reads them, checked.

Indexed by (schedule, contract_year), the year read by value, once, as MYGA_US_S’s lookup matches it. Projection.surr_charge_rate, the table’s one reader, reads its arrays, so a projection meets the check first. A contract year that is not a whole number, text and a blank among them, a blank schedule, a rate that is not a number from 0 to 1, and a schedule whose years do not run from 1, one row a year, without a gap or a repeat, are refused, naming the file, the column and the lines they are on, and so is a table that lacks initial or renewal, the two schedules Projection.surr_charge_id names: MYGA_US_S reads a schedule it cannot find as no charge, so one misspelt, Renewal, would leave every renewal period free of charge without a word. A year past a schedule’s last has no charge.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

surr_charge_age_caps()[source]#

The attained-age cap, as surr_charge_age_cap_table() reads it, checked.

Indexed by attained age, read by value, once, as MYGA_US_S’s lookup matches it. Projection.surr_charge_age_cap_at, the table’s one reader, reads its array, and takes its extents, surr_charge_age_cap_first() and surr_charge_age_cap_last(), from the same checked ages before it reads a cap, so a projection meets the check first. An age that is not a whole number, text and a blank among them, a negative age, a cap that is not a number from 0 to 1, and ages that do not run one row an age, without a gap or a repeat, are refused, naming the file, the column and the lines they are on. MYGA_US_S takes int() of the raw ages for its extents: one written 94x turns them to text, the extents then compare as text, and the cap vanishes. A negative age is refused because surr_charge_age_cap_array() holds the table from age 0, where MYGA_US_S projects a point on such a table.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

surr_charge_age_cap_first()[source]#

The first attained age surr_charge_age_caps() gives a cap for, an int.

surr_charge_age_cap_last()[source]#

The last attained age surr_charge_age_caps() gives a cap for, an int.

rate_scenario_steps()[source]#

The rate scenarios, as rate_scenario() reads them, checked.

Indexed by (scenario_id, t), each t a valuation month read by value, once, as MYGA_US_S’s step function compares it. Projection.scenario_rate, the table’s one reader, reads its arrays, so a projection meets the check first. A month that is not a whole number, text and a blank among them, a blank scenario_id, a yield or a market rate that is not a finite number, and a scenario that sets two levels at one month are refused, naming the file, the column and the lines they are on. MYGA_US_S’s step function compares the raw months with t: one written as text raises Python’s own TypeError, and a blank one is a level that never holds. A scenario may start after t = 0, for business issued later, so the month a scenario must start by is the point’s: model_point_table() refuses a point whose scenario sets no level by the first month the point is on the books, where the step function would find none and raise Python’s own ValueError.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

wd_schedules()[source]#

The withdrawal programmes, as withdrawal_table() reads them, checked.

Indexed by (wd_schedule_id, t), each t a policy month read by value, once, as MYGA_US_S’s exact-key lookup matches it. Projection.wd_scheduled_pp, the table’s one reader, reads its array, so a projection meets the check first. A month that is not a whole number, text and a blank among them, a blank wd_schedule_id, an amount that is not a number, zero or more, and a programme that withdraws twice in one month are refused, naming the file, the column and the lines they are on. MYGA_US_S’s lookup matches the raw months: one written as text turns them all to text, and every withdrawal of every programme is then zero.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

mva_factors()[source]#

The declared-differential duration factors, read and checked.

Indexed by whole years remaining, read by value, once, as MYGA_US_S’s lookup matches them. mva_factor_array(), which Projection.mva_duration_factor, the table’s one reader, reads, is built from it, and that cell takes the table’s last year, mva_factor_last_year(), from the same checked years before it reads a factor, so a projection meets the check first. A year that is not a whole number, text and a blank among them, a factor that is not a finite number, and years that do not run from 0, one row a year, without a gap or a repeat, are refused, naming the file, the column and the lines they are on. MYGA_US_S takes int() of the raw years for the last one: one written as text raises a KeyError that names no file.

A key of 2**53 or more in size is refused too, before it is read into int64, which would hold one past int64 as the least int64 without a word.

mva_factor_last_year()[source]#

The last whole year remaining mva_factors() gives a factor for, an int.

point_id()[source]#

The point_id of every model point, in table order: int64 [point].

Projection is indexed by position, idx, so this is how a projection names its point: Projection.point_id() is the element at idx, and every refusal quotes it. A label that is not a whole number is refused, naming every such label, and so is one of 2**53 or more in size, which numpy would read as the least int64 once past int64. pandas reads a column of whole labels as int64 exactly, so the bound refuses some labels the vector could hold, none of them a label a table needs.

age_at_entry()[source]#

The issue age (ANB) of every model point: int64 [point].

The age_at_entry column, which model_point_table() holds to a whole number of years below 2**53 in size, where the int64 holds MYGA_US_S’s int() of it.

duration_mth()[source]#

Each model point’s completed policy months at the valuation date: int64 [point].

The duration_mth column, which model_point_table() holds to a whole number of months below 2**53 in size.

guar_period()[source]#

The guarantee period in years of every model point: int64 [point].

The guar_period column, which model_point_table() holds to a whole number of years, 1 or more and below 2**53 in size.

policy_count()[source]#

The contracts each model point stands for: float64 [point].

The policy_count column, which model_point_table() holds to a number, zero or more.

model_point_numbers()[source]#

The model point table’s numeric columns, each value read by float(), checked.

MYGA_US_S reads each of these columns as float(model_point()[column]) at the first month a projection asks for it, so this reads every value so, element by element, once: the eight columns of the contract the projection reads as numbers, premium, declared_rate_initial, gmir, mgsv_rate, mgsv_annual_charge, free_wd_rate, mva_ref_yield_at_issue and premium_tax_rate, and the ten of the opening state, which model_point_table() has checked already. A value float() reads is read as it reads it: a blank, which pandas reads as NaN, whitespace alone included, is NaN and projects NaN, as there; an infinity is one; 1_000 is 1000.0. A value float() cannot read, text, is refused, naming the column and every point it is on, where MYGA_US_S raises ValueError at the first month that reads it. No range is imposed on the eight: MYGA_US_S imposes none. The vectors of the eighteen, premium() to mva_ref_yield_locked_init(), read this, so a projection meets the refusal at its first read of any of them.

premium()[source]#

P, the single purchase payment per contract: float64 [point].

The premium column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

declared_rate_initial()[source]#

The declared rate guaranteed for the initial term: float64 [point].

The declared_rate_initial column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

gmir()[source]#

The guaranteed minimum interest rate: float64 [point].

The gmir column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

mgsv_rate()[source]#

The contract GMSV accumulation rate: float64 [point].

The mgsv_rate column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

mgsv_annual_charge()[source]#

The annual contract charge deducted from the floor: float64 [point].

The mgsv_annual_charge column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

free_wd_rate()[source]#

The free-withdrawal percentage under pct_av: float64 [point].

The free_wd_rate column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

mva_ref_yield_at_issue()[source]#

The MVA reference yield locked at issue: float64 [point].

The mva_ref_yield_at_issue column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

premium_tax_rate()[source]#

The premium tax rate: float64 [point].

The premium_tax_rate column as model_point_numbers() reads it, float() of each value, as MYGA_US_S reads it: a blank is NaN.

av_pp_init()[source]#

The account value per contract entering month 0: float64 [point].

The av_pp_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

mgsv_pp_init()[source]#

The Model #805 floor per contract entering month 0: float64 [point].

The mgsv_pp_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

prem_accum_gmir_pp_init()[source]#

P_accum@GMIR per contract entering month 0: float64 [point].

The prem_accum_gmir_pp_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

interest_credited_pp_init()[source]#

The interest credited from issue to the valuation date: float64 [point].

The interest_credited_pp_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

interest_credited_pp_prev_anniv_init()[source]#

The interest credited to the anniversary opening the year under way: float64 [point].

The interest_credited_pp_prev_anniv_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

free_wd_allow_init()[source]#

The free-withdrawal allowance of the year under way: float64 [point].

The free_wd_allow_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

free_wd_remain_init()[source]#

The unused allowance carried into month 0: float64 [point].

The free_wd_remain_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

tax_basis_pp_init()[source]#

The IRC 72 basis entering month 0, or at issue: float64 [point].

The tax_basis_pp_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

credit_rate_init()[source]#

The declared rate in force entering month 0: float64 [point].

The credit_rate_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

mva_ref_yield_locked_init()[source]#

The MVA reference yield in force entering month 0: float64 [point].

The mva_ref_yield_locked_init column of the opening state, which model_point_table() holds to a finite number, as model_point_numbers() reads it.

is_wd_gross()[source]#

Whether each model point’s mgsv_wd_convention is gross: int64 [point].

1 where mgsv_wd_convention is "gross", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_wd_net_of_charges()[source]#

Each model point’s mgsv_wd_convention decoded, net_of_charges: int64 [point].

1 where mgsv_wd_convention is "net_of_charges", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_rollover()[source]#

Whether each model point’s renewal_architecture is rollover: int64 [point].

1 where renewal_architecture is "rollover", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_annual_redeclare()[source]#

renewal_architecture is annual_redeclare, by point: int64 [point].

1 where renewal_architecture is "annual_redeclare", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_pct_av()[source]#

Whether each model point’s free_wd_rule is pct_av: int64 [point].

1 where free_wd_rule is "pct_av", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_interest_only()[source]#

Whether each model point’s free_wd_rule is interest_only: int64 [point].

1 where free_wd_rule is "interest_only", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_linear_duration()[source]#

Whether each model point’s mva_family is linear_duration: int64 [point].

1 where mva_family is "linear_duration", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_geometric()[source]#

Whether each model point’s mva_family is geometric: int64 [point].

1 where mva_family is "geometric", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_declared_differential()[source]#

Each model point’s mva_family decoded, declared_differential: int64 [point].

1 where mva_family is "declared_differential", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_cap_none()[source]#

Whether each model point’s mva_cap_rule is none: int64 [point].

1 where mva_cap_rule is "none", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_cap_sym_sc()[source]#

Whether each model point’s mva_cap_rule is sym_sc: int64 [point].

1 where mva_cap_rule is "sym_sc", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_cap_min_sc_interest()[source]#

Whether each model point’s mva_cap_rule is min_sc_interest: int64 [point].

1 where mva_cap_rule is "min_sc_interest", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_cap_asym_sc_snfl()[source]#

Whether each model point’s mva_cap_rule is asym_sc_snfl: int64 [point].

1 where mva_cap_rule is "asym_sc_snfl", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

is_cap_gmir_floor()[source]#

Whether each model point’s mva_cap_rule is gmir_floor: int64 [point].

1 where mva_cap_rule is "gmir_floor", compared element by element and exactly, as MYGA_US_S compares it, and 0 elsewhere (the Space docstring, What cannot be decoded).

free_wd_mva_exempt()[source]#

Whether each point’s free allowance is exempt from charge and MVA: int64 [point].

1 where MYGA_US_S’s bool() of the free_wd_mva_exempt value is True, and 0 where it is False, element by element: True and False as pandas reads them, and also a blank, which pandas reads as NaN, and any text, both of which bool() reads as True. That is MYGA_US_S’s reading, kept (the Space docstring, What cannot be decoded).

scenario_count()[source]#

The number of scenarios rate_scenario.csv holds, the table checked: an int.

The rows of mva_ref_yield_array() and market_rate_array() are 0, which no point is numbered, and 1 to this.

scenario_number()[source]#

Each model point’s rate scenario, numbered from 1: int64 [point].

The scenarios of rate_scenario_steps() are numbered from 1 in the sorted order of their ids, and each point is given the number of its scenario_id, matched by value as MYGA_US_S’s .loc matches it, so an id written 1.0 is the scenario 1. model_point_table() has refused a point whose scenario the table lacks, so no point is numbered 0.

scenario_first_mths()[source]#

Each scenario’s first valuation month, 0 if before: int64 [scenario].

Element s is the first t of scenario s (scenario_number()), or 0 where that is negative, since the step function then holds a level from month 0; element 0 is 0. Projection.scenario_rate refuses a month before it.

scenario_last_mth()[source]#

The last valuation month the scenario arrays hold: an int.

The table’s last t, or 0 if it is negative, and read_last_mth at the most, the last month any read reaches (the Space docstring): the last column of mva_ref_yield_array() and market_rate_array(), from which on every scenario’s last level holds, as the step function holds it, up to the last month a read reaches.

mva_ref_yield_array()[source]#

The MVA reference yield by scenario and valuation month: float64 [scenario, t].

Element [s, m] is the level of mva_ref_yield that scenario s’s step function holds at month m, 0 to scenario_last_mth(): the level of the scenario’s latest row at or before m, a row before month 0 included, as MYGA_US_S’s step function finds it. NaN before the scenario’s first month, scenario_first_mths(), and in row 0, neither of which Projection.scenario_rate reads. A row past scenario_last_mth() sets no element: no read reaches its month.

market_rate_array()[source]#

The market competitor rate by scenario and valuation month: float64 [scenario, t].

The market_rate column, held as mva_ref_yield_array() holds its own.

wd_schedule_count()[source]#

The number of withdrawal programmes withdrawal_table.csv holds, checked: an int.

wd_schedule_number()[source]#

Each model point’s withdrawal programme, numbered from 1: int64 [point].

Numbered as scenario_number() numbers the scenarios, from the programmes of wd_schedules(). model_point_table() has refused a point whose programme the table lacks, so no point is numbered 0.

wd_last_mth()[source]#

The last policy month wd_amount_array() holds: an int.

The table’s last t, or 0 if it is negative, and read_last_mth at the most, past the last policy month in cover (the Space docstring). Projection.wd_scheduled_pp reads no withdrawal past it.

wd_amount_array()[source]#

The gross withdrawal by programme and policy month: float64 [schedule, month].

Element [s, m] is the wd_amount of programme s (wd_schedule_number()) at policy month m, 0 to wd_last_mth(), and 0.0 at a month the programme has no row: MYGA_US_S’s lookup reads a key that misses as no withdrawal, so this is a value, not padding. Row 0 is NaN, unread. A row at a negative policy month is left out: MYGA_US_S reads the table at duration_mth(t) in cover alone, 0 or more, so never reads one, and so is a row past wd_last_mth(), a month past cover.

surr_charge_last_years()[source]#

Each surrender charge schedule’s last contract year: int64 [3].

Element 1 is the initial schedule’s last year and element 2 the renewal schedule’s, the rows Projection.INITIAL and Projection.RENEWAL of surr_charge_array(); element 0, Projection.NONE, is 0. From the years surr_charge_scales() has checked, so each schedule runs from year 1.

surr_charge_array()[source]#

The surrender charge rate by schedule and contract year: float64 [3, year].

Row 1 is the initial schedule and row 2 the renewal schedule, by contract year from 1, the codes Projection.INITIAL and Projection.RENEWAL of the schedule Projection.surr_charge_id names. NaN in row 0, Projection.NONE, in column 0 and past each schedule’s last year, surr_charge_last_years(), none of which Projection.surr_charge_rate reads. Other schedules the table may hold are checked by surr_charge_scales(), and not held: the projection reads none.

surr_charge_age_cap_array()[source]#

The attained-age cap on the surrender charge by attained age: float64 [age].

Element a is the surr_charge_cap of age a in surr_charge_age_cap.csv, from 0 to the last age, surr_charge_age_cap_last(); NaN below the first, surr_charge_age_cap_first(), which Projection.surr_charge_age_cap_at never reads.

mort_schedule_count()[source]#

The number of sexes mort_table.csv rates, the table checked: an int.

mort_schedule()[source]#

Each model point’s sex, numbered from 1: int64 [point].

Numbered as scenario_number() numbers the scenarios, from the sexes of mort_rates() in sorted order. model_point_table() has refused a point whose sex the table lacks, so no point is numbered 0.

mort_age_firsts()[source]#

Each sex’s first attained age in mort_table.csv: int64 [schedule].

Element s is the first age of sex s (mort_schedule()), 0 or more since mort_rates() refuses a negative age; element 0 is 0.

mort_age_lasts()[source]#

Each sex’s last attained age in mort_table.csv: int64 [schedule].

As mort_age_firsts(), the last age. Projection.proj_len refuses a point whose sex’s last age is below the attained age of its last covered month, maturity_age - 1, a Projection Reference this Space cannot read.

mort_table_array()[source]#

The annual mortality rate by sex and attained age: float64 [schedule, age].

Element [s, a] is the mort_rate of sex s (mort_schedule()) at age a in mort_table.csv, from age 0 to the table’s last. NaN in row 0, below each sex’s first age and past its last, mort_age_firsts() and mort_age_lasts(), none of which Projection.mort_rate_at reads.

mva_factor_array()[source]#

The duration factors by whole year remaining: float64 [year, column].

Element [k, 0] is factor_lo, the Ic < 6% column, at k whole years remaining, and [k, 1] factor_hi, the Ic >= 6% one [S14], k from 0 to mva_factor_last_year(), every year mva_factors() has checked, so nothing is padding.