Files
daqbuffer/items/src/scalarevents.rs

511 lines
13 KiB
Rust

use crate::minmaxavgbins::MinMaxAvgBins;
use crate::numops::NumOps;
use crate::streams::{Collectable, Collector};
use crate::{
ts_offs_from_abs, Appendable, ByteEstimate, Clearable, EventAppendable, FilterFittingInside, Fits, FitsInside,
PushableIndex, RangeOverlapInfo, ReadPbv, ReadableFromFile, SitemtyFrameType, TimeBinnableType,
TimeBinnableTypeAggregator, WithLen, WithTimestamps,
};
use err::Error;
use netpod::log::*;
use netpod::NanoRange;
use serde::{Deserialize, Serialize};
use std::fmt;
use tokio::fs::File;
// TODO in this module reduce clones.
// TODO add pulse.
#[derive(Serialize, Deserialize)]
pub struct ScalarEvents<NTY> {
pub tss: Vec<u64>,
pub values: Vec<NTY>,
}
impl<NTY> SitemtyFrameType for ScalarEvents<NTY>
where
NTY: NumOps,
{
const FRAME_TYPE_ID: u32 = crate::EVENT_VALUES_FRAME_TYPE_ID + NTY::SUB;
}
impl<NTY> ScalarEvents<NTY> {
pub fn empty() -> Self {
Self {
tss: vec![],
values: vec![],
}
}
}
impl<NTY> fmt::Debug for ScalarEvents<NTY>
where
NTY: fmt::Debug,
{
fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
write!(
fmt,
"count {} ts {:?} .. {:?} vals {:?} .. {:?}",
self.tss.len(),
self.tss.first(),
self.tss.last(),
self.values.first(),
self.values.last(),
)
}
}
impl<NTY> WithLen for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn len(&self) -> usize {
self.tss.len()
}
}
impl<NTY> WithTimestamps for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn ts(&self, ix: usize) -> u64 {
self.tss[ix]
}
}
impl<NTY> ByteEstimate for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn byte_estimate(&self) -> u64 {
if self.tss.len() == 0 {
0
} else {
// TODO improve via a const fn on NTY
self.tss.len() as u64 * 16
}
}
}
impl<NTY> RangeOverlapInfo for ScalarEvents<NTY> {
fn ends_before(&self, range: NanoRange) -> bool {
match self.tss.last() {
Some(&ts) => ts < range.beg,
None => true,
}
}
fn ends_after(&self, range: NanoRange) -> bool {
match self.tss.last() {
Some(&ts) => ts >= range.end,
None => panic!(),
}
}
fn starts_after(&self, range: NanoRange) -> bool {
match self.tss.first() {
Some(&ts) => ts >= range.end,
None => panic!(),
}
}
}
impl<NTY> FitsInside for ScalarEvents<NTY> {
fn fits_inside(&self, range: NanoRange) -> Fits {
if self.tss.is_empty() {
Fits::Empty
} else {
let t1 = *self.tss.first().unwrap();
let t2 = *self.tss.last().unwrap();
if t2 < range.beg {
Fits::Lower
} else if t1 > range.end {
Fits::Greater
} else if t1 < range.beg && t2 > range.end {
Fits::PartlyLowerAndGreater
} else if t1 < range.beg {
Fits::PartlyLower
} else if t2 > range.end {
Fits::PartlyGreater
} else {
Fits::Inside
}
}
}
}
impl<NTY> FilterFittingInside for ScalarEvents<NTY> {
fn filter_fitting_inside(self, fit_range: NanoRange) -> Option<Self> {
match self.fits_inside(fit_range) {
Fits::Inside | Fits::PartlyGreater | Fits::PartlyLower | Fits::PartlyLowerAndGreater => Some(self),
_ => None,
}
}
}
impl<NTY> PushableIndex for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn push_index(&mut self, src: &Self, ix: usize) {
self.tss.push(src.tss[ix]);
self.values.push(src.values[ix].clone());
}
}
impl<NTY> Appendable for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn empty_like_self(&self) -> Self {
Self::empty()
}
fn append(&mut self, src: &Self) {
self.tss.extend_from_slice(&src.tss);
self.values.extend_from_slice(&src.values);
}
}
impl<NTY> Clearable for ScalarEvents<NTY> {
fn clear(&mut self) {
self.tss.clear();
self.values.clear();
}
}
impl<NTY> ReadableFromFile for ScalarEvents<NTY>
where
NTY: NumOps,
{
fn read_from_file(_file: File) -> Result<ReadPbv<Self>, Error> {
// TODO refactor types such that this can be removed.
panic!()
}
fn from_buf(_buf: &[u8]) -> Result<Self, Error> {
panic!()
}
}
impl<NTY> TimeBinnableType for ScalarEvents<NTY>
where
NTY: NumOps,
{
type Output = MinMaxAvgBins<NTY>;
type Aggregator = EventValuesAggregator<NTY>;
fn aggregator(range: NanoRange, x_bin_count: usize, do_time_weight: bool) -> Self::Aggregator {
debug!(
"TimeBinnableType for EventValues aggregator() range {:?} x_bin_count {} do_time_weight {}",
range, x_bin_count, do_time_weight
);
Self::Aggregator::new(range, do_time_weight)
}
}
pub struct EventValuesCollector<NTY> {
vals: ScalarEvents<NTY>,
range_complete: bool,
timed_out: bool,
}
impl<NTY> EventValuesCollector<NTY> {
pub fn new() -> Self {
Self {
vals: ScalarEvents::empty(),
range_complete: false,
timed_out: false,
}
}
}
impl<NTY> WithLen for EventValuesCollector<NTY> {
fn len(&self) -> usize {
self.vals.tss.len()
}
}
#[derive(Serialize)]
pub struct EventValuesCollectorOutput<NTY> {
#[serde(rename = "tsAnchor")]
ts_anchor_sec: u64,
#[serde(rename = "tsMs")]
ts_off_ms: Vec<u64>,
#[serde(rename = "tsNs")]
ts_off_ns: Vec<u64>,
values: Vec<NTY>,
#[serde(skip_serializing_if = "crate::bool_is_false", rename = "finalisedRange")]
range_complete: bool,
#[serde(skip_serializing_if = "crate::bool_is_false", rename = "timedOut")]
timed_out: bool,
}
impl<NTY> Collector for EventValuesCollector<NTY>
where
NTY: NumOps,
{
type Input = ScalarEvents<NTY>;
type Output = EventValuesCollectorOutput<NTY>;
fn ingest(&mut self, src: &Self::Input) {
self.vals.append(src);
}
fn set_range_complete(&mut self) {
self.range_complete = true;
}
fn set_timed_out(&mut self) {
self.timed_out = true;
}
fn result(self) -> Result<Self::Output, Error> {
let tst = ts_offs_from_abs(&self.vals.tss);
let ret = Self::Output {
ts_anchor_sec: tst.0,
ts_off_ms: tst.1,
ts_off_ns: tst.2,
values: self.vals.values,
range_complete: self.range_complete,
timed_out: self.timed_out,
};
Ok(ret)
}
}
impl<NTY> Collectable for ScalarEvents<NTY>
where
NTY: NumOps,
{
type Collector = EventValuesCollector<NTY>;
fn new_collector(_bin_count_exp: u32) -> Self::Collector {
Self::Collector::new()
}
}
pub struct EventValuesAggregator<NTY> {
range: NanoRange,
count: u64,
min: Option<NTY>,
max: Option<NTY>,
sumc: u64,
sum: f32,
int_ts: u64,
last_ts: u64,
last_val: Option<NTY>,
do_time_weight: bool,
}
impl<NTY> EventValuesAggregator<NTY>
where
NTY: NumOps,
{
pub fn new(range: NanoRange, do_time_weight: bool) -> Self {
let int_ts = range.beg;
Self {
range,
count: 0,
min: None,
max: None,
sum: 0f32,
sumc: 0,
int_ts,
last_ts: 0,
last_val: None,
do_time_weight,
}
}
// TODO reduce clone.. optimize via more traits to factor the trade-offs?
fn apply_min_max(&mut self, val: NTY) {
self.min = match &self.min {
None => Some(val.clone()),
Some(min) => {
if &val < min {
Some(val.clone())
} else {
Some(min.clone())
}
}
};
self.max = match &self.max {
None => Some(val),
Some(max) => {
if &val > max {
Some(val)
} else {
Some(max.clone())
}
}
};
}
fn apply_event_unweight(&mut self, val: NTY) {
let vf = val.as_prim_f32();
self.apply_min_max(val);
if vf.is_nan() {
} else {
self.sum += vf;
self.sumc += 1;
}
}
fn apply_event_time_weight(&mut self, ts: u64) {
if let Some(v) = &self.last_val {
let vf = v.as_prim_f32();
let v2 = v.clone();
self.apply_min_max(v2);
let w = if self.do_time_weight {
(ts - self.int_ts) as f32 * 1e-9
} else {
1.
};
if vf.is_nan() {
} else {
self.sum += vf * w;
self.sumc += 1;
}
self.int_ts = ts;
} else {
debug!(
"apply_event_time_weight NO VALUE {}",
ts as i64 - self.range.beg as i64
);
}
}
fn ingest_unweight(&mut self, item: &<Self as TimeBinnableTypeAggregator>::Input) {
for i1 in 0..item.tss.len() {
let ts = item.tss[i1];
let val = item.values[i1].clone();
if ts < self.range.beg {
} else if ts >= self.range.end {
} else {
self.apply_event_unweight(val);
self.count += 1;
}
}
}
fn ingest_time_weight(&mut self, item: &<Self as TimeBinnableTypeAggregator>::Input) {
for i1 in 0..item.tss.len() {
let ts = item.tss[i1];
let val = item.values[i1].clone();
if ts < self.int_ts {
debug!("just set int_ts");
self.last_ts = ts;
self.last_val = Some(val);
} else if ts >= self.range.end {
debug!("after range");
return;
} else {
debug!("regular");
self.apply_event_time_weight(ts);
self.count += 1;
self.last_ts = ts;
self.last_val = Some(val);
}
}
}
fn result_reset_unweight(&mut self, range: NanoRange, _expand: bool) -> MinMaxAvgBins<NTY> {
let avg = if self.sumc == 0 {
None
} else {
Some(self.sum / self.sumc as f32)
};
let ret = MinMaxAvgBins {
ts1s: vec![self.range.beg],
ts2s: vec![self.range.end],
counts: vec![self.count],
mins: vec![self.min.clone()],
maxs: vec![self.max.clone()],
avgs: vec![avg],
};
self.int_ts = range.beg;
self.range = range;
self.count = 0;
self.min = None;
self.max = None;
self.sum = 0f32;
self.sumc = 0;
ret
}
fn result_reset_time_weight(&mut self, range: NanoRange, expand: bool) -> MinMaxAvgBins<NTY> {
// TODO check callsite for correct expand status.
if true || expand {
debug!("result_reset_time_weight calls apply_event_time_weight");
self.apply_event_time_weight(self.range.end);
} else {
debug!("result_reset_time_weight NO EXPAND");
}
let avg = {
let sc = self.range.delta() as f32 * 1e-9;
Some(self.sum / sc)
};
let ret = MinMaxAvgBins {
ts1s: vec![self.range.beg],
ts2s: vec![self.range.end],
counts: vec![self.count],
mins: vec![self.min.clone()],
maxs: vec![self.max.clone()],
avgs: vec![avg],
};
self.int_ts = range.beg;
self.range = range;
self.count = 0;
self.min = None;
self.max = None;
self.sum = 0f32;
self.sumc = 0;
ret
}
}
impl<NTY> TimeBinnableTypeAggregator for EventValuesAggregator<NTY>
where
NTY: NumOps,
{
type Input = ScalarEvents<NTY>;
type Output = MinMaxAvgBins<NTY>;
fn range(&self) -> &NanoRange {
&self.range
}
fn ingest(&mut self, item: &Self::Input) {
debug!("ingest len {}", item.len());
if self.do_time_weight {
self.ingest_time_weight(item)
} else {
self.ingest_unweight(item)
}
}
fn result_reset(&mut self, range: NanoRange, expand: bool) -> Self::Output {
debug!("Produce for {:?} next {:?}", self.range, range);
if self.do_time_weight {
self.result_reset_time_weight(range, expand)
} else {
self.result_reset_unweight(range, expand)
}
}
}
impl<NTY> EventAppendable for ScalarEvents<NTY>
where
NTY: NumOps,
{
type Value = NTY;
fn append_event(ret: Option<Self>, ts: u64, value: Self::Value) -> Self {
let mut ret = if let Some(ret) = ret { ret } else { Self::empty() };
ret.tss.push(ts);
ret.values.push(value);
ret
}
}