The thread: The model has parameters — page 3
The filter rebuilt with every write
A log-structured store chooses a range filter's two prefix lengths when it writes a level, and a filling store's small levels hold only the keys that arrived last. Choosing at every write was predicted to keep the lead a pair chosen once on the full store has, except where keys arrive a cluster at a time and a small level's choice goes stale when the next batch lands elsewhere. It cannot go stale: a levelled store changes a level's keys only by rewriting it, and rewriting it rebuilds the filter. Choosing at every write and choosing once then differ by at most 2.3% of what a perfect filter saves, on both arrival orders, at two bits a key and at ten, and choosing costs twenty-eight times the hashing. What the order of arrival changes is the store: filled a cluster at a time, its small levels hold one cluster each, and at two bits a key the same filter captures 94% of a perfect filter's saving against 68% shuffled.
What is taught wronglyThe cap that binds on one text and not another
A periodic text looks like the one made of chains and pays 1.11 times the phrases for a cap of four. A text that repeats itself pays 2.96. The guess is backwards, and the reason is that depth measures nesting rather than repetition.
What is taught wronglyA sixth of what, exactly
The saving from a counting-only reverse half is 30.5% at one sampled position in four and 14.0% at one in a hundred and twenty-eight. A number quoted without its sampling rate is a number about a setting somebody chose.
Two parametersA watch that looks less
A landmark watch that refits when an estimated saving pays loses to the plain corners when traffic jumps every thousand queries, because its looking costs more than its refits save. Using its learned horizon to look less often there — doubling the interval when the horizon falls below half of it, halving it when the horizon comes back — was predicted to tie the corners at that pace and cost nothing elsewhere. It ties nothing: 150.5 cells a query against the corners' 142.7, since the looks it skips are the looks that found the chances. And at one jump in 2,500, where the watch had won, it loses by twenty cells, because the rule cannot come back down. Halving needs a horizon as long as the stretched interval, and at 8,000 queries no stream here supplies one. A two-state rule that can come back does better and still loses at both paces. The looking was never the waste; at one jump in 1,000 there is nothing to watch for.
When it does not fitOne bit a key, divided
A levelled store of a million clustered keys, each level's range filter a pair of prefix lengths chosen when the level is written, captures 68% of a perfect filter's saving at two bits a key spread evenly when its keys arrive shuffled, and 94% when they arrive a cluster at a time. The proposal was to divide the memory by what each level's filter saves per bit, and it predicted the memory would leave the small levels, whose keys the arrival order had already sorted out. It goes the other way, as it did for single prefix lengths, and much further: the small levels take twelve to sixteen bits a key, the deepest level one, and the store captures 99% of a perfect filter's saving on both orders while leaving part of the budget unspent. Divided, one bit a key on average does what four bits spread evenly cannot.
The data that is not a numberThe optimum the flag took away
A bidirectional index's wavelet-tree levels were coded in blocks of 63 bits because a sweep found 63 the best length: shorter blocks pay for more classes, longer ones for offsets that approach the block's own length. Then the classes were coded as runs and each offset given a flag to choose a code of its value changes instead. Lengthening the blocks to 127 bits was predicted to save on English words and the repetitive collection and to change nothing on random letters. It saves on all four texts, and there is no best length short of 1,023 bits on three of them: English words fall from 1.431 bits a character at 63 to 1.304 at 1,023. Under the codings the optimum came from, the same levels cost 3.728 at 1,023. The flag turned a long block's offset into the cost of its change points, which no block length moves, and left the length to be set by the one thing still growing with it: the work of a rank.
When it does not fitA division chosen before the queries arrive
A levelled store's range-filter memory, divided among its levels by trying every assignment against the store's own queries, captures 99% of a perfect filter's saving on both arrival orders, with the small levels at twelve to sixteen bits a key and the deepest at one. A store writing a level cannot try every assignment, but it knows its levels' sizes, and a rule giving each level δ more bits a key for every factor of four it is smaller was predicted to match the fitted division on the shuffled store and to land near the even split's 94% on the store filled a cluster at a time. It matches the fitted division on both: 97.2% to 99.0% shuffled and 99.7% to 99.9% filled a cluster at a time from one bit a key up, with δ chosen on a sample of queries and charged on others. One δ of four, set before any query, is within a point and a half everywhere from one bit a key up. The rule fails exactly where the proposal said: a δ of six or more at one bit a key gives the deepest level nothing, and the shuffled store falls from 97.2% to 73.9%.
Two parametersA watch that switches itself off
A landmark watch that cannot beat the plain corners by looking less was given a switch instead: at every look it prices the corners on the same twenty queries, and when the corners would have been cheaper over its last four looks it goes back to them and stops looking for 8,000 queries. At one jump in 1,000 it works — 145.5 cells a query comparing search costs, 144.4 charging the watch for its own looking and refits, against the corners' 142.7 and the watch's 150.7. At one jump in 2,500, where watching pays, both switches give the watch's advantage away, 125.3 and 126.1 against 116.4, and the charging switch is off for 60% of the stream, as much as at the fast pace. What a look reads is the same at both paces: 12,249 cells a window at one, 11,257 at the other, with a spread of sixty thousand. The watch's value arrives in a few large looks after jumps, and four looks at any pace are mostly the others.
The data that is not a numberA flag that weighs the work as well as the bits
A wavelet-tree level's mixed blocks are each coded by their arrangement's index or by the gaps between their value changes, a flag choosing the shorter, and longer blocks then save bits on every text at the price of more work a rank. Charging the flag for decoding work as well as bits was predicted to reach the same curve as shortening the block, and to beat it by no more than a hundredth of a bit a character. On English words, DNA and protein it reaches nothing below that curve. On the repetitive collection it reaches 0.156 bits a character below it: 1,023-bit blocks with a price on work cost 0.458 bits a character for less work a rank than 63-bit blocks do at 0.607. The difference is in which blocks the flag codes by their arrangement — on English words blocks that change value ninety-six times, on the repetitive collection blocks holding one short run of ones in two hundred and fifty-five bits.
Two parametersA switch that counts the jumps
A landmark watch whose switch read its own costs could not tell fast traffic from slow, because what a look reads has the same mean at both paces. A switch that reads the traffic instead — the mean spacing of the last ten times the destination moved more than ten cells — was predicted to reach the corners at one jump in 1,000 and keep the watch at every slower pace. It separates the paces where costs could not: 146.6 cells a query at one jump in 1,000 against the watch's 150.7, and on drifting traffic it never switches off at all. But a count of k events is a reading with a delay of k events and a noise of one part in √k, and no k has both: three stretch ends reach the corners' 142.7 at the fast pace and are off 38% of the slower stream; twenty never misfire and save nothing. And when traffic slows, the events that would show it stop arriving, so the count keeps its fast reading and the switch stays off: 135.8 cells a query on a stream that slows halfway, worse than the corners or the watch.
When it does not fitThe sizes a store passes through
A levelled store's range-filter memory can be divided by its levels' sizes alone, a few more bits a key for every factor of four a level is smaller than the deepest, and on the full store that rule matched the division fitted to the store's own queries at every budget. Through the fill it does not. At a bit a key or less it falls up to 17 points of a perfect filter's saving behind on the shuffled store and 26 on the store filled a cluster at a time, and every large loss has the same cause: part way through the fill the deepest level is a smaller share of the store, and the rule gives it nothing.
What is taught wronglyThe occurrences a join invents
Eight documents run together hold forty-nine eight-character windows that span a join, twenty-three of which occur in no document at all. Every index built over the concatenation reports them, and five essays of this collection paid that cost silently.
The data that is not a numberA price the levels share
A bidirectional index's wavelet-tree levels code each mixed block by its arrangement or by its gaps, and a flag that charges each code for the work a rank costs inside it reached below the curve the block lengths draw on one text of four. Setting that price separately for each level of the tree was predicted to find a few hundredths of a bit on English words, where the deep levels hold runs. It finds nothing on any text: under a ten-thousandth of a bit beyond one shared price. The price a block wants is set by the text, not by the depth of the node it sits in, and every level holds a fifth of the index.
When it does not fitA floor under every level
A levelled store's filter memory divided by its levels' sizes lost up to 26 points of a perfect filter's saving part way through its fill, every large loss a level given no filter at all. A floor of half a bit a key under every level was predicted to close every gap above five points at one bit a key and cost nothing at two and four. It closes them: the worst gap at one bit falls from 17 points to 4 on the shuffled store and from 25 to under 1 on the clustered one, and nothing moves at two bits and up. At half a bit a key the floor is the whole budget and the rule becomes the even split, which the clustered store wants half way through and the shuffled store, at the same sizes, loses 35 points by.
The data that is not a numberThe blocks the searches ask for
A flag that codes each block of an index's wavelet-tree levels as an arrangement or a gap code, charged for the work a rank does inside the block, was priced as if ranks landed on every position equally, and a price for each level found nothing that one price did not. Pattern searches do not ask evenly: on English words the busiest tenth of the blocks takes 22% of their ranks, and 35% when common patterns are asked more. Priced by the work of the ranks that actually land in each block, chosen on one sample of searches and charged on another, the flag reaches 0.11 to 0.16 bits a character below the old pricing at the same work on English, and 0.03 on the repetitive collection.
When it does not fitA span the queries do not see
A rule for dividing a levelled store's range-filter memory by its levels' sizes, with a floor of half a bit a key, came within five points of the division fitted to the store's own queries at every budget from one bit up — but at half a bit the shuffled and the clustered store wanted opposite divisions of identically sized levels. A level's fences, its first and last keys, say how much of the key space it spans, and a clustered level spans little. Scaling each level's bits by that span was predicted to close the half-bit gap and change nothing above. It changes nothing at half a bit — where the floor is the whole budget, so there is nothing left to scale — and from one bit up it puts the clustered store 30 points behind the fitted division, where the floored rule was within one. A level holding one cluster spans 0.4% of the key space and meets 4.4% of the queries, because half of them start at a stored key: the fences measure the key space, and the queries are aimed at the keys.
When it does not fitA level scaled by the queries it meets
A levelled store's range-filter memory, divided among its levels by their sizes with a floor of half a bit a key, comes within a few points of the division fitted to its own queries from one bit a key up; at half a bit the shuffled and the clustered store want opposite divisions. Letting the floor step aside below one bit was predicted to fix the shuffled store and cost the clustered one, and it does exactly that: 9.3 points behind the fitted division on the shuffled store instead of 34.8, 26.1 on the clustered instead of 9.1. Scaling each level by the share of queries its fences can meet — the measure the span rule should have read — brings the shuffled store to 2.2 points, but on that store every level meets the same 91% of queries, and the plain size rule with its slope cut by a tenth does the same. On the clustered store the same scaling is the worst rule measured, 41.5 points behind: its newest level meets 3% of queries and is the level the fitted division protects.
What a bound isTwo at binary, five at twenty-six
The saving is a factor in the alphabet, so a two-symbol alphabet gets two. Approximate matching in this field is mostly done on DNA, which sits near the bottom of the list at 2.7.
What is taught wronglyThe dial that has no setting
The generator has one parameter. The real version history's run count asks it for 2.3% and its phrase count asks for 3.2% — and the reason is not that the dial is badly calibrated. Real edits average 7.3 characters a block and generated ones average 1.04.
What a bound isThe flat bottom of a shallow curve
The low width is chosen as the floor of log of the universe over the count. Rounding it up instead costs one bit on five thousand, because the total is m·w plus n over two to the w and the minimum is where those two are equal.
What a bound isFlat in the budget, and not
One saving is eleven times at every error budget, because it is a property of the alphabet. The other moves between ninety-eight and a hundred and five, because it follows the share of extensions that find nothing. Two savings, two shapes, and neither line crosses the other.
What a bound isThe exponential is in the expression
The subset construction on one family reaches two to the k plus one states, exactly and not approximately. A literal of the same length gives eleven. Both are regular expressions and the difference is that one of them asks the machine to remember something.
What the libraries doA looser budget wastes a larger share
More errors permitted means more work, and the fraction of that work which was never going to help rises with it — from thirty-one per cent at no errors to seventy-five at two. The saving is worth most where the search is most expensive.
StructuresBits and steps on one frame
The size falls from ninety-nine per cent to eighty-six as the sampling thins, and the walk to a sampled position rises from two and a half steps to sixty-four. Neither line is the answer; the answer is a point on the pair.
What the libraries doWhere the table starts paying
Five thousand and sixty-five operations before the first character, then one per character. Against nothing before the first character and thirty-nine per character. They cross at two hundred and fifty-six characters, and that crossing is what an engine's compile decision actually is.
What a bound isA recut the law decides
Parts of a file cut by the square-root law where selection queries fall lost their advantage once the queries moved a percentile and a half. A partition that recuts itself when the law says the saving pays for the rewrite was predicted to stay within 20% of the best fixed recut interval at every drift speed, and to fail on traffic that jumps. With recent queries decayed at a half-life of 250, it stays within 7% up to a drift of 0.1 percentiles every hundred queries and within 22% at five times that. On jumping traffic it beats every interval fixed in advance, rewriting 13 times a stream where the best interval rewrites 80. The law does not remove a setting: it trades the interval, which has to move fourfold with the speed, for a half-life that does not.
What a bound isEvery cut the law moves
A file cut into seven parts by the square-root law, and recut when the law says the saving pays, was charged a full rewrite for every recut. Charging only the parts whose boundaries change was predicted to cut a recut's cost to a tenth on slow drift. It cuts it by 2%. When the band of hot queries moves, the law moves 98% of its cuts, a median of nineteen thousand ranks each, because the parts on either side of the band have to share out what the band left. Holding small moves back keeps a quarter of the file and costs more a query; recutting only the parts around the band costs up to three times as much.
What a bound isThe two parts the law was not given
The square-root law moves almost every cut of a seven-part file whenever its band of hot queries moves. A background cut once by keys, with only the part holding the band cut again, was proposed instead: a recut would rewrite a seventh of the file. It does, and it beats the law at every drift speed — 365 transfers a query against 417 at a moderate drift — but it wins with nine parts to the law's seven. Given nine, the law costs 322. Held to seven, the fixed background loses to the law at every speed. What survives the fair comparison is narrower and still useful: six parts by keys, with the band's part cut in four, ties the nine-part law on drift for a quarter of its rewriting. It loses on jumps, because a fixed background waits about three hundred queries to believe the band has left before it moves, and it pays double wherever a background cut runs through the band.
What a bound isA background that holds two bands
A file kept in six fixed parts by keys, with the part holding the band of hot queries cut in four, ties the nine-part square-root law on drifting traffic and loses to it by 10% to 16% when the band jumps, because after a jump it pays for a whole background part for three hundred queries before moving its split. Asking every part whether a split would pay, instead of asking which part holds the most queries, was proposed to end the wait. It changes nothing: 363 transfers a query against 365. What ends the wait is the second half of the proposal, letting two parts be split at once in three and two pieces: on jumps the background comes within 2% to 4% of the law, and on fast drift, which nobody expected it to touch, it beats the law by 6% to 10%, because a band lying across a background cut is held by two split parts rather than by one split and one whole.
What the libraries doPieces that fit the buffer
A sorted file with a late tenth appended costs the buffer-limited stable merge sort 2.93 moves a key at a buffer of a 64th, fifteen times the least any sort could move, and two thirds of it is the tail sorting itself. Taking the tail a buffer at a time, inserting each key into a sorted piece in the buffer and merging the piece into the head, was predicted to bring that under two moves a key at small buffers. It never comes within twice the merge sort: insertion costs a quarter of a piece a key, and the smaller the pieces the more often each one's merge crosses the head's top fifth. The same pieces sorted in place by merging do pay, at 2.41 moves a key, and for a reason the proposal did not name: a piece that fits the buffer merges into the head without a single rotation. The gain lasts only while the tail belongs near the top of the file.