Curve-path colouring and cache mode settings are used (and can thus be
factored out of) all child types; this moves them into the parent type's
`.__init__()` and adjusts all sub-types match:
- the bulk was moved out of the `Curve.__init__()` including all
previous commentary around cache settings.
- adjust `BarItems` to use a `NoCache` mode and instead use the
`last_step_pen: pg.Pen` and `._pen` inside it's `.pain()` instead of
defining functionally duplicate vars.
- adjust all (transitive) calls to `BarItems` to use the new kwargs
names.
Allows callers to know if they should care about a particular viz
rendering call by immediately knowing if the graphics are in view. This
turns out super useful particularly when doing dynamic y-ranging overlay
calcs.
First, we rename what was `chart_maxmin()` -> `multi_maxmin()` and don't
`partial` it in to the `DisplayState`, just call it with correct `Viz`
ref inputs.
Second, as we've done with `ChartView.maybe_downsample_graphics()` use
the output from the main `Viz.update_graphics()` and feed it to the
`.maxmin()` calls for the ohlc and vlm chart but still deliver the same
output signature as prior. Also accept and use an optional profiler
input, drop `DisplayState.maxmin()` and add `.vlm_viz`.
Further perf related tweak to do with more efficient incremental
updates:
- only call `multi_maxmin()` if the main fast chart viz does a pixel
column step.
- mask out hist viz and vlm viz and all linked fsp `._set_yrange()`
calls for now until we figure out how to best optimize these updates
when considering the new group-scaled-by-% style for multicharts.
- drop `.enable_auto_yrange()` calls during startup.
Acts as short cut when pipe-lining from `Viz.update_graphics()` (which
now returns the needed in-view array-relative-read-slice as output) such
that `Viz.read()` and `.datums_range()` doesn't need to be called
internally multiple times. In this case where `i_read_range` is provided
we of course skip doing time index translations and consequently lookup
the appropriate (epoch-time) index indices for caching.
Removes the multi-maxmin usage as well as ensures appropriate `Viz` refs
are passed into the view methods now requiring it. Also drops the "back
linking" of the vlm chart view to the source OHLC chart since we're
going to add this as a default to the charting API.
The max min for a given data range is defined on the lowest level
through the `Viz` api intermingling it with the view is a layering
issue. Instead make `._set_yrange()` call the appropriate view's viz
(since they should be one-to-one) directly and thus avoid any callback
monkey patching nonsense.
Requires that we now make `._set_yrange()` require either one of an
explicit `yrange: tuple[float, float]` min/max pair or the `Viz` ref (so
that maxmin can be called) as input. Adjust
`enable/disable_auto_yrange()` to bind in a new `._yranger()` partial
that's (solely) needed for signal reg/unreg which binds in the now
required input `Viz` to these methods.
Comment the `autoscale_overlays` block in `.maybe_downsample_graphics()`
for now until we figure out the most sane way to auto-range all linked
overlays and subplots (with their own overlays).
In an effort to ensure uniform and uppx-optimized last datum graphics
updates call this method directly instead of the equivalent graphics
object thus ensuring we only update the last pixel column according with
the appropriate max/min computed from the last uppx's worth of data.
Fixes / improvements to enable `.draw_last()` usage include,
- change `Viz._render_table` -> `._alt_r: tuple[Renderer, pg.GraphicsItem] | None`
which holds an alternative (usually downsampled) render and graphics
obj.
- extend the `.draw_last()` signature to include:
- `last_read` to allow passing in the already read data from
`.update_graphics()`, if it isn't passed then a manual read is done
internally.
- `reset_cache: bool` which is passed through to the graphics obj.
- use the new `Formatter.flat_index_ratio: float` when indexing into xy
1d data to compute the max/min for that px column.
Other,
- drop `bars_range` input from `maxmin()` since it's unused.
For the purposes of avoiding another full format call we can stash the
last rendered 1d xy pre-graphics formats as
`IncrementalFormatter.x/y_1d: np.ndarray`s and allow readers in the viz
and render machinery to use this data easily for things like "only
drawing the last uppx's worth of data as a line". Also add
a `.flat_index_ratio: float` which can be used similarly as a scalar
applied to indexes into the src array but instead when indexing
(flattened) 1d xy formatted outputs. Finally, this drops the way
overdone/noisy `.__repr__()` meth we had XD
When a new tick comes in but no new time step / bar is yet needed (to be
appended) we can simply adjust **only** the last bar datum
lines-graphic(s) to avoid a redraw of the preceding `QPainterPath` on
every tick. Do this by calling `Viz.draw_last()` on the fast and slow
chart and adjusting the guards around calls to `Viz.update_graphics()`
(which *does* update paths) to only enter when there's a `do_px_step`
condition. We can stop calling `main_viz.plot.vb._set_yrange()` on view
treading cases since the range should have already been adjusted by the
clearing-tick processing mxmn updates.
Further this changes,
- the `chart_maxmin()` helper (which we should eventually just get rid
of) to take bound in `Viz`s for the ohlc and vlm chart instead of the
chart widget handles.
- extend the guard around hist viz yranging to only enter when not in
"axis mode" - the same as for the fast viz.
Since we removed the `Viz.update_graphics()` call from the main rt loop
we have to be sure to call it in the history chart incr-loop to avoid
a gap between the last bar and prior history since startup. We only
need to update on tread since that should be the only time a full redraw
is ever necessary, ow only the last datum is needed.
Further this moves the graphics cycle func's profiler init to the top in
an effort to get more correct latency measures.
Since `ChartPlotWidget.update_graphics_from_flow()` is more or less just
a call to `Viz.update_graphics()` try to call that directly where
possible.
Changes include:
- calling the viz in the display state specific `maxmin()`.
- passing a viz instance to each `ChartView._set_yrange()` call (in prep
of explicit group auto-ranging); not that this input is unused in the
method for now.
- drop `bars_range` var passing since we don't use it.
Inside `._interaction` routines we need access to `Viz` instances.
Instead of doing `CharPlotWidget._vizs: dict` lookups this ensures each
plot can lookup it's (parent) viz without error.
Also, adjusts `Viz.maxmin()` output parsing to new signature.
Move the `Viz.datums_range()` call into `Viz.maxmin()` itself thus
minimizing the chart `.maxmin()` method to an ultra light wrapper around
the viz call. Also move all profiling into the `Viz` method.
Adjust `Viz.maxmin()` to return both the (rounded) x-range values which
correspond to the range containing the y-domain min and max so that
it can be used for up and coming overlay group maxmin calcs.
We obviously don't want to be debugging a sample-index issue if/when the
market for the asset is closed (since we'll be guaranteed to have
a mismatch, lul). Pass in the `feed_is_live: trio.Event` throughout the
backfilling routines to allow first checking for the live feed being active
so as to avoid breakpointing on false +ves. Also, add a detailed warning
log message for when *actually* investigating a mismatch.
This should never really happen but when it does it appears to be a race
with writing startup pre-graphics-formatter array data where we get
`x_end` epoch value subtracting some really small offset value (like
`-/+0.5`) or the opposite where the `x_start` is epoch and `x_end` is
small.
This adds a warning msg and `breakpoint()` as well as guards around the
entire code downsampling code path so that when resumed the downsampling
cycle should just be skipped and avoid a crash.
This attempt was unsuccessful since trying to (re)select the last
highlighted item on both an "enter" or "click" of that item causes
a hang and then segfault in `Qt`; no clue why..
Adds a `keep_current_item_selected: bool` flag to
`CompleterView.show_cache_entries()` but using it seems to always cause
a hang and crash; we keep all potential use spots commented for now
obviously to avoid this. Also included is a bunch of tidying to logic
blocks in the kb-control loop for readability.
Whenever the last datum is in view `slice_from_time()` need to always
spec the final array index (i.e. the len - 1 value we set as
`read_i_max`) to avoid a uniform-step arithmetic error where gaps in the
underlying time series causes an index that's too low to be returned.
- adjust zoom focal to be min of the view-right coord or the right-most
point on the flow graphic in view and drop all the legacy l1-in-view
focal point cruft.
- flip to not auto-scaling overlays by default.
- change the `._set_yrange()` margin to `0.09`.
- drop `use_vr: bool` usage.
Before this axes were being stacked from the outside in (for `'right'`
and 'bottom'` axes) which is somewhat non-intuitive for an `.append()`
operation. As such this change makes a symbol list stack a set of
`'right'` axes from left-to-right.
Details:
- rename `ComposeGridLayout.items` -> `.pitems`
- return `(int, list[AxisItem])` pairs from `.insert/append_plotitem()`
and the down stream `PlotItemOverlay.add_plotitem()`.
- drop `PlotItemOverlay.overlays` and add it back as `@property` around
the underlying `.layout.pitems`.
Factor and fix dst <- src pair parsing into a new func
`get_likely_pair()` and use throughout initial position loading; solves
a parsing bug for src asset balances which aren't only 3 chars long..
a terrible assumption.
Goes back to always adjusting the y-axis range to include the L1 spread
and clearing label in view whenever the last datum is also in view,
previously this was broken after reworking the display loop for
multi-feeds.
Drops a bunch of old commented tick looping cruft from before we started
using tick-type framing. Also adds more stringent guards for ignoring
but error logging quote values that are more then 25% out of range; it
seems particularly our `ib` feed has some issues with strange `price`
values that are way off here and there?
Instead of having the l1 lines be inside the view space, move them to be
inside their respective axis (with only a 16 unit portion inside the
view) such that the clear price label can overlay with them nicely
without obscuring; this is much better suited to multiple adjacent
y-axes and in general is simpler and less noisy.
Further `L1Labels` + `LevelLabel` style tweaks:
- adjust `.rect` positioning to be "right" (i.e. inside the parent
y-axis) with a slight 16 unit shift toward the viewbox (using the new
`._x_br_offset`) to allow seeing each level label's line even when the
clearing price label is positioned at that same level.
- add a newline's worth of vertical space to each of the bid/ask labels
so that L1 labels' text content isn't ever obscured by the clear price
label.
- set a low (10) z-value to ensure l1 labels are always placed
underneath the clear price label.
- always fill the label rect with the chosen background color.
- make labels fully opaque so as to always make them hide the parent
axes' `.tickStrings()` contents.
- make default color the "default" from the global scheme.
- drop the "price" part from the l1 label text contents, just show the
book-queue's amount (in dst asset's units, aka the potential clearing vlm).
In the case where the last-datum-graphic hasn't been created yet, simply
return a `None` from this method so the caller can choose to ignore the
output. Further, drop `.px_width()` since it makes more sense defined on
`Viz` as well as the previously commented `BarItems.x_uppx()` method.
Also, don't round the `.x_uppx()` output since it can then be used when
< 1 to do x-domain scaling during high zoom usage.
Factor some common methods into the parent type:
- `.x_uppx()` for reading the horizontal units-per-pixel.
- `.x_last()` for reading the "closest to y-axis" last datum coordinate
for zooming "around" during mouse interaction.
- `.px_width()` for computing the max width of any curve in view in
pixels.
Adjust all previous derived `pg.GraphicsObject` child types to now
inherit from this new parent and in particular enable proper `.x_uppx()`
support to `BarItems`.
Use proper uppx scaling when either of scaling the data to the x-domain
index-range or when the uppx is < 1 (now that we support it) such that
both the fast and slow chart always appropriately scale and offset to
the y-axis with the last datum graphic just adjacent to the order line
arrow markers.
Further this fixes the `.index_step()` calc to use the "earliest" 16
values to compute the expected sample step diff since the last set often
contained gaps due to start up race conditions and generated
unexpected/incorrect output.
Further this drops the `.curve_width_pxs()` method and replaces it with
`.px_width()`, taken from the graphics object API and instead returns
the pixel account for the whole view width instead of the
x-domain-data-range within the view.
Doesn't seem like we really need to handle the situation where the start
or stop input time stamps are outside the index range of the data since
the new binary search handling via `numpy.searchsorted()` covers this
case at minimal runtime cost and with an equally correct output. Allows
us to drop some other indexing endpoint internal variables as well.
We want the fast and slow chart to behave the same on calls to
`Viz.default_view()` so adjust the offset calc to make both work:
- just offset by the line len regardless of step / uppx
- add back the `should_line: bool` output from `render_bar_items()` (and
use it to set a new `ds_allowed: bool` guard variable) so that we can
bypass calling the m4 downsampler unless the bars have been switched
to the interpolation line graphic (which we normally required before
any downsampling of OHLC graphics data).
Further, this drops use of the `use_vr: bool` flag from all rendering
since we pretty much always use it by default.
Previously with array-int indexing we had to map the input x-domain
"indexes" passed to `DynamicDateAxis._indexes_to_timestr()`. In the
epoch-time indexing case we obviously don't need to lookup time stamps
from the underlying shm array and can instead just cast to `int` and
relay the values verbatim.
Further, this patch includes some style adjustments to `AxisLabel` to
better enable multi-feed chart overlays by avoiding L1 label clutter
when multiple y-axes are stacked adjacent:
- adjust the `Axis` typical max string to include a couple spaces suffix
providing for a bit more margin between side-by-side y-axes.
- make the default label (fill) color the "default" from the global
color scheme and drop it's opacity to .9
- add some new label placement options and use them in the
`.boundingRect()` method:
* `._x/y_br_offset` for relatively shifting the overall label relative
to it's parent axis.
* `._y_txt_h_scaling` for increasing the bounding rect's height
without including more whitespace in the label's text content.
- ensure labels have a high z-value such that by default they are always
placed "on top" such that when we adjust the l1 labels they can be set
to a lower value and thus never obscure the last-price label.
Turns out we were updating the wrong ``Viz``/``DisplayState`` inside the
closure style `increment_history_view()`` (probably due to looping
through the flumes and dynamically closing in that task-func).. Instead
define the history incrementer at module level and pass in the
`DisplayState` explicitly. Further rework the `DisplayState` attrs to be
more focused around the `Viz` associated with the fast and slow chart
and be sure to adjust output from each `Viz.incr_info()` call to latest
update. Oh, and just tweaked the line palette for the moment.
FYI "treading" here is referring to the x-shifting of the curve when
the last datum is in view such that on new sampled appends the "last"
datum is kept in the same x-location in UI terms.
Mainly it was the global (should we )increment logic that needs to be
independent for the fast vs. slow chart such that the slow isn't
update-shifted by the fast and vice versa. We do this using a new
`'i_last_slow'` key in the `DisplayState.globalz: dict` which is
singleton for each sample-rate-specific chart and works for both time
and array indexing.
Also, we drop some old commented `graphics.draw_last_datum()` code that
never ended up being needed again inside the coordinate cache reset
bloc.
Might as well since it makes the chart look less gappy and we can easily
flip the index switch now B)
Also adds a new `'i_slow_last'` key to `DisplayState` for a singleton
across all slow charts and thus no more need for special case logic in
`viz.incr_info()`.
Define the x-domain coords "offset" (determining the curve graphics
per-datum placement) for each formatter such that there's only on place
to change it when needed. Obviously each graphics type has it's own
dimensionality and this is reflected by the array shapes on each
subtype.
Previously we were drawing with the middle of the bar on each index with
arms to either side: +/- some arm length. Instead this changes so that
each bar is drawn *after* each index/timestamp such that in graphics
coords the bar span more correctly matches the time span in the
x-domain. This makes the linked region between slow and fast chart
directly match (without any transform) for epoch-time indexing such that
the last x-coord in view on the fast chart is no more then the
next time step in (downsampled) slow view.
Deats:
- adjust in `._pathops.path_arrays_from_ohlc()` and take an `bar_w` bar
width input (normally taken from the data step size).
- change `.ui._ohlc.bar_from_ohlc_row()` and
`BarItems.draw_last_datum()` to match.
Allows easily switching between normal array `int` indexing and time
indexing by just flipping the `Viz._index_field: str`.
Also, guard all the x-data audit breakpoints with a time indexing
condition.
Turned out to be super simple to get the first draft to work since the
fast and slow chart now use the same domain, however, it seems like
maybe there's an offset issue still where the fast may be a couple
minutes ahead of the slow?
Need to dig in a bit..
Using a global "last index step" (via module var) obviously has problems
when working with multiple feed sets in a single global app instance:
any separate feed-set will be incremented according to an app-global
index-step and thus won't correctly calc per-feed-set-step update info.
Impl deatz:
- drop `DisplayState.incr_info()` (since previously moved to `Viz`) and
call that method on each appropriate `Viz` instance where necessary;
further ensure the appropriate `DisplayState` instance is passed in to
each call and make sure to pass a `state: DisplayState`.
- add `DisplayState.hist_vars: dict` for history chart (sets) to
determine the per-feed (not set) current slow chart (time) step.
- add `DisplayState.globalz: dict` to house a common per-feed-set state
and use it inside the new `Viz.incr_info()` such that
a `should_increment: bool` can be returned and used by the display
loop to determine whether to x-shift the current chart.
Read the `Viz.index_step()` directly to avoid always reading 1 on the
slow chart; this was completely broken before and resulting in not
rendering the bars graphic on the slow chart until at a true uppx of
1 which obviously doesn't work for 60 width bars XD
Further cleanups to `._render` module:
- drop `array` output from `Renderer.render()`, `read_from_key` input
and fix type annot.
- drop `should_line`, `changed_to_line` and `render_kwargs` from
`render_baritems()` outputs and instead calc `should_redraw` logic
inside the func body and return as output.
First allocation vs. first "prepend" of source data to an xy `ndarray`
format **must be mutex** in order to avoid a double prepend.
Previously when both blocks were executed we'd end up with
a `.xy_nd_start` that was decremented (at least) twice as much as it
should be on the first `.format_to_1d()` call which is obviously
incorrect (and causes problems for m4 downsampling as discussed below).
Further, since the underlying `ShmArray` buffer indexing is managed
(i.e. write-updated) completely independently from the incremental
formatter updates and internal xy indexing, we can't use
`ShmArray._first.value` and instead need to use the particular `.diff()`
output's prepend length value to decrement the `.xy_nd_start` on updates
after initial alloc.
Problems this resolves with m4:
- m4 uses a x-domain diff to calculate the number of "frames" to
downsample to, this is normally based on the ratio of pixel columns on
screen vs. the size of the input xy data.
- previously using an int-index (not epoch time) the max diff between
first and last index would be the size of the input buffer and thus
would never cause a large mem allocation issue (though it may have
been inefficient in terms of needed size).
- with an epoch time index this max diff could explode if you had some
near-now epoch time stamp **minus** an x-allocation value: generally
some value in `[0.5, -0.5]` which would result in a massive frames and
thus internal `np.ndarray()` allocation causing either a crash in
`numba` code or actual system mem over allocation.
Further, put in some more x value checks that trigger breakpoints if we
detect values that caused this issue - we'll remove em after this has
been tested enough.
Turns out we can't seem to avoid the artefacts when click-drag-scrolling
(results in weird repeated "smeared" curve segments) so just go back to
the original code.
Ensures that a "last datum" graphics object exists so that zooming can
read it using `.x_last()`. Also, disable the linked region stuff for now
since it's totally borked after flipping to the time indexing.
Since we don't really need it defined on the "chart widget" move it to
a viz method and rework it to hell:
- always discard the invalid view l > r case.
- use the graphic's UPPX to determine UI-to-scene coordinate scaling for
the L1-label collision detection, if there is no L1 just offset by
a few (index step scaled) datums; this allows us to drop the 2x
x-range calls as was hacked previous.
- handle no-data-in-view cases explicitly and error if we get any
ostensibly impossible cases.
- expect caller to trigger a graphics cycle if needed.
Further support this includes a rework a slew of other important
details:
- add `Viz.index_step`, an idempotent computed, index (presumably uniform)
step value which is needed for variable sample rate graphics displayed
on an epoch (second) time index.
- rework `Viz.datums_range()` to pass view x-endpoints as first and last
elements in return `tuple`; tighten up snap-to-data edge case logic
using `max()`/`min()` calls and better internal var naming.
- adjust all calls to `slice_from_time()` to not expect an "abs" slice.
- drop all `.yrange` resetting since we can just have the `Renderer` do
it when necessary.
If we presume that time indexing using a uniform step we can calculate
the exact index (using `//`) for the input time presuming the data
set has zero gaps. This gives a massive speedup over `numpy` fancy
indexing and (naive) `numba` iteration. Further in the case where time
gaps are detected, we can use `numpy.searchsorted()` to binary search
for the nearest expected index at lower latency.
Deatz,
- comment-disable the call to the naive `numba` scan impl.
- add a optional `step: int` input (calced if not provided).
- add todos for caching binary search results in the gap detection
cases.
- drop returning the "absolute buffer indexing" slice since the caller
can always just use the read-relative slice to acquire it.
When we use an epoch index and any sample rate > 1s we need to scale the
"number of bars" to that step in order to place the view correctly in
x-domain terms. For now we're calcing the step in-method but likely,
longer run, we'll pull this from elsewhere (like a ``Viz`` attr).
Gives approx a 3-4x speedup using plain old iterate-with-for-loop style
though still not really happy with this .5 to 1 ms latency..
Move the core `@njit` part to a `_slice_from_time()` with a pure python
func with orig name around it. Also, drop the output `mask` array since
we can generally just use the slices in the caller to accomplish the
same input array slicing, duh..
We need to subtract the first index in the array segment read, not the
first index value in the time-sliced output, to get the correct offset
into the non-absolute (`ShmArray.array` read) array..
Further we **do** need the `&` between the advance indexing conditions
and this adds profiling to see that it is indeed real slow (like 20ms
ish even when using `np.where()`).
Again, to make epoch indexing a flip-of-switch for testing look up the
`Viz.index_field: str` value when updating labels.
Also, drops the legacy tick-type set tracking which we no longer use
thanks to the new throttler subsys and it's framing msgs.
Planning to put the formatters into a new mod and aggregate all path
gen/op helpers into this module.
Further tweak include:
- moving `path_arrays_from_ohlc()` back to module level
- slice out the last xy datum for `OHLCBarsAsCurveFmtr` 1d formatting
- always copy the new x-value from the source to `.x_nd`
This was a major cause of error (particularly trying to get epoch
indexing working) and really isn't necessary; instead just have
`.diff()` always read from the underlying source array for current
index-step diffing and append/prepend slice construction.
Allows us to,
- drop `._last_read` state management and thus usage.
- better handle startup indexing by setting `.xy_nd_start/stop` to
`None` initially so that the first update can be done in one large
prepend.
- better understand and document the step curve "slice back to previous
level" logic which is now heavily commented B)
- drop all the `slice_to_head` stuff from and instead allow each
formatter to choose it's 1d segmenting.
In an effort to make it easy to override the indexing scheme.
Further, this repairs the `.datums_range()` special case to handle when
the view box is to-the-right-of the data set (i.e. l > datum_start).
As in make the call to `Flume.slice_from_time()` to try and convert any
time index values from the view range to array-indices; all untested
atm.
Also drop some old/unused/moved methods:
- `._set_xlimits()`
- `.bars_range()`
- `.curve_width_pxs()`
and fix some `flow` -> `viz` var naming.
Don't expect values (array + slice) to be returned and applied by
`.incr_update_xy_nd()` and instead presume this will implemented
internally in each (sub)formatter.
Attempt to simplify some incr-update routines, (particularly in the step
curve formatter, though most of it was reverted to just a simpler form
of the original implementation XD) including:
- dropping the need for the `slice_to_head: int` control.
- using the `xy_nd_start/stop` index counters over custom lookups.
Remove harcoded `'index'` field refs from all formatters in a first
attempt at moving towards epoch-time alignment (though don't actually
use it it yet).
Adjustments to the formatter interface:
- property for `.xy_nd` the x/y nd arrays.
- property for and `.xy_slice` the nd format array(s) start->stop index
slice.
Internal routine tweaks:
- drop `read_src_from_key` and always pass full source array on updates
and adjust handlers to expect to have to index the data field of
interest.
- set `.last_read` right after update calls instead of after 1d
conversion.
- drop `slice_to_head` array read slicing.
- add some debug points for testing 'time' indexing (though not used
here yet).
- add `.x_nd` array update logic for when the `.index_field` is not
'index' - i.e. when we begin to try and support epoch time.
- simplify some new y_nd updates to not require use of `np.broadcast()`
where possible.
Probably means it doesn't need to be a `Flume` method but it's
convenient to expect the caller to pass in the `np.ndarray` with
a `'time'` field instead of a `timeframe: str` arg; also, return the
slice mask instead of the sliced array as output (again allowing the
caller to do any slicing). Also, handle the slice-outside-time-range
case by just returning the entire index range with a `None` mask.
Adjust `Viz.view_data()` to instead do timeframe (for rt vs. hist shm
array) lookup and equiv array slicing with the returned mask.
Since these modules no longer contain Qt specific code we might
as well include them in the data sub-package.
Also, add `IncrementalFormatter.index_field` as single point to def the
indexing field that should be used for all x-domain graphics-data
rendering.
Was broken since the `_adhoc_futes_set` rework a while back. Removes the
cmdty symbols from that set into a new one and fixes the contract
case block to catch `Contract(secType='CMDTY')` case. Also makes
`Client.search_symbols()` return details `dict`s so that `piker search`
will work again..
Since higher level charting and fsp management need access to the
new `Flume` indexing apis this adjusts some func sigs to pass through
(and/or create) flume instances:
- `LinkedSplits.add_plot()` and dependents.
- `ChartPlotWidget.draw_curve()` and deps, and it now returns a `Flow`.
- `.ui._fsp.open_fsp_admin()` and `FspAdmin.open_fsp_ui()` related
methods => now we wrap the destination fsp shm in a flume on the admin
side and is returned from `.start_engine_method()`.
Drop a bunch of (unused) chart widget methods including some already
moved to flume methods: `.get_index()`, `.in_view()`,
`.last_bar_in_view()`, `.is_valid_index()`.
Move to expect and process new by-tick-event frames where the display
loop can now just iterate the most recent tick events by type instead of
the entire tick history sequence - thus we reduce iterations inside the
update loop.
Also, go back to use using the detected display's refresh rate (minus 6)
as the default feed requested throttle rate since we can now handle
much more bursty-ness in display updates thanks to the new framing
format B)
Factor out the chart widget creation since it's only executed once
during rendering of the first feed/flow whilst keeping plotitem overlay
creation inside the (flume oriented) init loop. Only create one vlm and
FSP chart/chain for now until we figure out if we want FSPs overlayed by
default or selected based on the "front" symbol in use. Add a default
color-palette set using shades of gray when plotting overlays. Presume
that the display loop's quote throttle rate should be uniformly
distributed over all input symbol-feeds for now. Restore feed pausing on
mouse interaction.
Initial support for real-time multi-symbol overlay charts using an
aggregate feed delivered by `Feed.open_multi_stream()`.
The setup steps for constructing the overlayed plot items is still very
very rough and will likely provide incentive for better refactoring high
level "charting APIs". For each fqsn passed into `display_symbol_data()`
we now synchronously,
- create a single call to `LinkedSplits.plot_ohlc_main() -> `ChartPlotWidget`
where we cache the chart in scope and for all other "sibling" fqsns
we,
- make a call to `ChartPlotWidget.overlay_plotitem()` -> `PlotItem`, hide its axes,
make another call with this plotitem input to
`ChartPlotWidget.draw_curve()`, set a sym-specific view box auto-yrange maxmin callback,
register the plotitem in a global `pis: dict[str, list[pgo.PlotItem, pgo.PlotItem]] = {}`
Once all plots have been created we then asynchronously for each symbol,
- maybe create a volume chart and register it in a similar task-global
table: `vlms: dict[str, ChartPlotWidget] = {}`
- start fsp displays for each symbol
Then common entrypoints are entered once for all symbols:
- a single `graphics_update_loop()` loop-task is started wherein
real-time graphics update components for each symbol are created,
* `L1Labels`
* y-axis last clearing price stickies
* `maxmin()` auto-ranger
* `DisplayState` (stored in a table `dss: dict[str, DisplayState] = {}`)
* an `increment_history_view()` task
and a single call to `Feed.open_multi_stream()` is used to create
a symbol-multiplexed quote stream which drives a single loop over all
symbols wherein for each quote the appropriate components are looked
up and passed to `graphics_update_cycle()`.
- a single call to `open_order_mode()` is made with the first symbol
provided as input, though eventually we want to support passing in the
entire list.
Further internal implementation details:
- special tweaks to the `pg.LinearRegionItem` setup wherein the region
is added with a zero opacity and *after* all plotitem overlays to
avoid and issue where overlays weren't being shown within the region
area in the history chart.
- all symbol-specific graphics oriented update calls are adjusted to
pass in the fqsn:
* `update_fsp_chart()`
* `ChartView._set_yrange()`
* ChartPlotWidget.update_graphics_from_flow()`
- avoid a double increment on sample step updates by not calling the
increment on any vlm chart since it seems the vlm-ohlc chart linking
already takes care of this now?
- use global counters for the last epoch time step to avoid incrementing
all views more then once per new time step given underlying shm array
buffers may be on different array-index values from one another.
Main "public" API change is to make `GodWidget.get/set_chart_symbol()`
accept and cache-on fqsn tuples to allow handling overlayed chart groups
and adjust method names to be plural to match.
Wrt `LinkedSplits`,
- create all chart widget axes with a `None` plotitem argument and set
the `.pi` field after axis creation (since apparently we have another
object reference causality dilemma..)
- set a monkeyed `PlotItem.chart_widget` for use in axes that still need
the widget reference.
- drop feed pause/resume for now since it's leaking feed tasks on the
`brokerd` side and we probably don't really need it any more, and if
we still do it should be done on the feed not the flume.
Wrt `ChartPlotItem`,
- drop `._add_sticky()` and use the `Axis` method instead and add some
overlay + axis sanity checks.
- refactor `.draw_ohlc()` to be a lighter wrapper around a call to
`.add_plot()`.
We have this method on our `ChartPlotWidget` but it makes more sense to
directly associate axis-labels with, well, the label's parent axis XD.
We add `._stickies: dict[str, YAxisLabel]` to replace
`ChartPlotWidget._ysticks` and pass in the `pg.PlotItem` to each axis
instance, stored as `Axis.pi` instead of handing around linked split
references (which are way out of scope for a single axis).
More work needs to be done to remove dependence on `.chart:
ChartPlotWidget` references in the date axis type as per comments.
Comments out the pixel-cache resetting since it doesn't seem we need it
any more to avoid draw oddities?
For `.fast_path` appends, this nearly got it working except the new path
segments are either not being connected correctly (step curve) or not
being drawn in full since the history path (plain line).
Leaving the attempted code commented in for a retry in the future; my
best guesses are that maybe,
- `.connectPath()` call is being done with incorrect segment length
and/or start point.
- the "appended" data: `appended = array[-append_len-1:slice_to_head]`
(done inside the formatter) isn't correct (i.e. endpoint handling
considering a path append) and needs special handling for different
curve types?
Ensure `.boundingRect()` calcs and `.draw_last_datum()` do geo-sizing
based on source data instead of presuming some `1.0` unit steps in some
spots; we need this to support an epoch index as is needed for overlays.
Further, clean out a bunch of old bounding rect calc code and add some
commented code for trying out `QRectF.united()` on the path + last datum
curve segment. Turns out that approach is slower as per eyeballing the
added profiler points.
After trying to hack epoch indexed time series and failing miserably,
decided to properly factor out all formatting routines into a common
subsystem API: ``IncrementalFormatter`` which provides the interface for
incrementally updating and tracking pre-path-graphics formatted data.
Previously this functionality was mangled into our `Renderer` (which
also does the work of `QPath` generation and update) but splitting it
out also preps for being able to do graphics-buffer downsampling and
caching on a remote host B)
The ``IncrementalFormatter`` (parent type) has the default behaviour of
tracking a single field-array on some source `ShmArray`, updating
a flattened `numpy.ndarray` in-mem allocation, and providing a default
1d conversion for pre-downsampling and path generation.
Changed out of `Renderer`,
- `.allocate_xy()`, `update_xy()` and `format_xy()` all are moved to
more explicitly named formatter methods.
- all `.x/y_data` nd array management and update
- "last view range" tracking
- `.last_read`, `.diff()`
- now calls `IncrementalFormatter.format_to_1d()` inside `.render()`
The new API gets,
- `.diff()`, `.last_read`
- all view range diff tracking through `.track_inview_range()`.
- better nd format array names: `.x/y_nd`, `xy_nd_start/stop`.
- `.format_to_1d()` which renders pre-path formatted arrays ready for
both m4 sampling and path gen.
- better explicit overloadable formatting method names:
* `.allocate_xy()` -> `.allocate_xy_nd()`
* `.update_xy()` -> `.incr_update_xy_nd()`
* `.format_xy()` -> `.format_xy_nd_to_1d()`
Finally this implements per-graphics-type formatters which define
each set up related formatting routines:
- `OHLCBarsFmtr`: std multi-line style bars
- `OHLCBarsAsCurveFmtr`: draws an interpolated line for ohlc sampled data
- `StepCurveFmtr`: handles vlm style curves
More or less a revamp (and possibly first draft for something similar in
`tractor` core) which ensures all actor trees attempt to discover the
`pikerd` registry actor.
Implementation improvements include:
- new `Registry` singleton which houses the `pikerd` discovery
socket-address `Registry.addr` + a `open_registry()` manager which
provides bootstrapped actor-local access.
- refine `open_piker_runtime()` to do the work of opening a root actor
and call the new `open_registry()` depending on whether a runtime has
yet been bootstrapped.
- rejig `[maybe_]open_pikerd()` in terms of the above.
Allows running simultaneous data feed services on the same (linux) host
by avoiding file-name collisions instead keying shm buffer sets by the
given `brokerd` instance. This allows, for example, either multiple dev
versions of the data layer to run side-by-side or for the test suite to
be seamlessly run alongside a production instance.
Previously we were relying on implicit actor termination in
`maybe_spawn_daemon()` but really on `pikerd` teardown we should be sure
to tear down not only all service tasks in each actor but also the actor
runtimes. This adjusts `Services.cancel_service()` to only cancel the
service task scope and wait on the `complete` event and reworks the
`open_context_in_task()` inner closure body to,
- always cancel the service actor at exit.
- not call `.cancel_service()` (potentially causing recursion issues on
cancellation).
- allocate a `complete: trio.Event` to signal full task + actor termination.
- pop the service task from the `.service_tasks` registry.
Further, add a `maybe_set_global_registry_sockaddr()` helper-cm to do
the work of checking whether a registry socket needs-to/has-been set
and use it for discovery calls to the `pikerd` service tree.
Seems that by default their history indexing rounds down/back to the
previous time step, so make sure we add a minute inside `Client.bars()`
when the `end_dt=None`, indicating "get the latest bar". Add
a breakpoint block that should trigger whenever the latest bar vs. the
latest epoch time is mismatched; we'll remove this after some testing
verifying the history bars issue is resolved.
Further this drops the legacy `backfill_bars()` endpoint which has been
deprecated and unused for a while.
Always use `open_sample_stream()` to register fast and slow quote feed
buffers and get a sampler stream which we use to trigger
`Sampler.broadcast_all()` calls on the service side after backfill
events.
Now spawned under the `pikerd` tree as a singleton-daemon-actor we offer
a slew of new routines in support of this micro-service:
- `maybe_open_samplerd()` and `spawn_samplerd()` which provide the
`._daemon.Services` integration to conduct service spawning.
- `open_sample_stream()` which is a client-side endpoint which does all
the work of (lazily) starting the `samplerd` service (if dne) and
registers shm buffers for update as well as connect a sample-index
stream for iterator by the caller.
- `register_with_sampler()` which is the `samplerd`-side service task
endpoint implementing all the shm buffer and index-stream registry
details as well as logic to ensure a lone service task runs
`Services.increment_ohlc_buffer()`; it increments at the shortest period
registered which, for now, is the default 1s duration.
Further impl notes:
- fixes to `Services.broadcast()` to ensure broken streams get discarded
gracefully.
- we use a `pikerd` side singleton mutex `trio.Lock()` to ensure
one-and-only-one `samplerd` is ever spawned per `pikerd` actor tree.
Drop the `_services` module level ref and adjust all client code to
match. Drop struct inheritance and convert all methods to class level.
Move `Brokerd.locks` -> `Services.locks` and add sampling mod to pikerd
enabled set.
We're moving toward a single actor managing sampler work and distributed
independently of `brokerd` services such that a user can run samplers on
different hosts then real-time data feed infra. Most of the
implementation details include aggregating `.data._sampling` routines
into a new `Sampler` singleton type.
Move the following methods to class methods:
- `.increment_ohlc_buffer()` to allow a single task to increment all
registered shm buffers.
- `.broadcast()` for IPC relay to all registered clients/shms.
Further add a new `maybe_open_global_sampler()` which allocates
a service nursery and assigns it to the `Sampler.service_nursery`; this
is prep for putting the step incrementer in a singleton service task
higher up the data-layer actor tree.
When we see multiple history frames that are duplicate to the request
set, bail re-trying after a number of tries (6 just cuz) and return
early from the tsdb backfill loop; presume that this many duplicates
means we've hit the beginning of history. Use a `collections.Counter`
for the duplicate counts. Make sure and warn log in such cases.
Wow, turns out tick framing was totally borked since we weren't framing
on "greater then throttle period long waits" XD
This moves all the framing logic into a common func and calls it in
every case:
- every (normal) "pre throttle period expires" quote receive
- each "no new quote before throttle period expires" (slow case)
- each "no clearing tick yet received" / only burst on clears case
Add some (untested) data slicing util methods for mapping time ranges to
source data indices:
- `.get_index()` which maps a single input epoch time to an equiv array
(int) index.
- add `slice_from_time()` which returns a view of the shm data from an
input epoch range presuming the underlying struct array contains
a `'time'` field with epoch stamps.
- `.view_data()` which slices out the "in view" data according to the
current state of the passed in `pg.PlotItem`'s view box.
This has been an outstanding idea for a while and changes the framing
format of tick events into a `dict[str, list[dict]]` wherein for each
tick "type" (eg. 'bid', 'ask', 'trade', 'asize'..etc) we create an FIFO
ordered `list` of events (data) and then pack this table into each
(throttled) send. This gives an additional implied downsample reduction
(in terms of iteration on the consumer side) from `N` tick-events to
a (max) `T` tick-types presuming the rx side only needs the latest tick
event.
Drop the `types: set` and adjust clearing event test to use the new
`ticks_by_type` map's keys.
Instead of uniformly distributing the msg send rate for a given
aggregate subscription, choose to be more bursty around clearing ticks
so as to avoid saturating the consumer with L1 book updates and vs.
delivering real trade data as-fast-as-possible.
Presuming the consumer is in the "UI land of slow" (eg. modern display
frame rates) such an approach serves more useful for seeing "material
changes" in the market as-bursty-as-possible (i.e. more short lived fast
changes in last clearing price vs. many slower changes in the bid-ask
spread queues). Such an approach also lends better to multi-feed
overlays which in aggregate tend to scale linearly with the number of
feeds/overlays; centralization of bursty arrival rates allows for
a higher overall throttle rate if used cleverly with framing.
Seems that by default their history indexing rounds down/back to the
previous time step, so make sure we add a minute inside `Client.bars()`
when the `end_dt=None`, indicating "get the latest bar". Add
a breakpoint block that should trigger whenever the latest bar vs. the
latest epoch time is mismatched; we'll remove this after some testing
verifying the history bars issue is resolved.
Further this drops the legacy `backfill_bars()` endpoint which has been
deprecated and unused for a while.
Likely pertains to helping with stuff in issues #345 and #373 and just
generally is handy to have when processing ledgers / clearing event
tables.
Adds the following helper methods:
- `iter_by_dt()` to iter-sort an arbitrary `Transaction`-like table of
clear entries.
- `Position.iter_clears()` as a convenience wrapper for the above.
Trying to send a message in the `NoBsWs.fixture()` exit when the ws is
not currently disconnected causes a double `._stack.close()` call which
will corrupt `trio`'s coro stack. Instead only do the unsub if we detect
the ws is still up.
Also drops the legacy `backfill_bars()` module endpoint.
Fixes#437
Seems that by default their history indexing rounds down/back to the
previous time step, so make sure we add a minute inside `Client.bars()`
when the `end_dt=None`, indicating "get the latest bar". Add
a breakpoint block that should trigger whenever the latest bar vs. the
latest epoch time is mismatched; we'll remove this after some testing
verifying the history bars issue is resolved.
Further this drops the legacy `backfill_bars()` endpoint which has been
deprecated and unused for a while.
Instead of requiring any `-b` try to import all built-in broker backend
python modules by default and only load those detected from the input symbol
list's fqsn values. In other words the `piker chart` cmd can be run sin
`-b` now and that flag is only required if you only want to load
a subset of the built-ins or are trying to load a specific
not-yet-builtin backend.
Allows using `set` ops for subscription management and guarantees no
duplicates per `brokerd` actor. New API is simpler for dynamic
pause/resume changes per `Feed`:
- `_FeedsBus.add_subs()`, `.get_subs()`, `.remove_subs()` all accept multi-sub
`set` inputs.
- `Feed.pause()` / `.resume()` encapsulates management of *only* sending
a msg on each unique underlying IPC msg stream.
Use new api in sampler task.
Previously we would only detect overruns and drop subscriptions on
non-throttled feed subs, however you can get the same issue with
a wrapping throttler task:
- the intermediate mem chan can be blocked either by the throttler task
being too slow, in which case we still want to warn about it
- the stream's IPC channel actually breaks and we still want to drop
the connection and subscription so it doesn't be come a source of
stale backpressure.
Set each quote-stream by matching the provider for each `Flume` and thus
results in some flumes mapping to the same (multiplexed) stream.
Monkey-patch the equivalent `tractor.MsgStream._ctx: tractor.Context` on
each broadcast-receiver subscription to allow use by feed bus methods as
well as other internals which need to reference IPC channel/portal info.
Start a `_FeedsBus` subscription management API:
- add `.get_subs()` which returns the list of tuples registered for the
given key (normally the fqsn).
- add `.remove_sub()` which allows removing by key and tuple value and
provides encapsulation for sampler task(s) which deal with dropped
connections/subscribers.
Adds provider-list-filtered (quote) stream multiplexing support allowing
for merged real-time `tractor.MsgStream`s using an `@acm` interface.
Behind the scenes we are just doing a classic multi-task push to common
mem chan approach.
Details to make it work on `Feed`:
- add `Feed.mods: dict[str, Moduletype]` and
`Feed.portals[ModuleType, tractor.Portal]` which are both populated
during init in `open_feed()`
- drop `Feed.portal` and `Feed.name`
Also fix a final lingering tsdb history loading loop termination bug.
A slight facepalm but, the main issue was a simple indexing logic error:
we need to slice with `tsdb_history[-shm._first.value:]` to push most
recent history not oldest.. This allows cleanup of tsdb backfill loop as
well.
Further, greatly simply `diff_history()` time slicing by using the
classic `numpy` conditional slice on the epoch field.
This had a bug prior where the end of a frame (a partial) wasn't being
sliced correctly and we'd get odd gaps showing up in the backfilled from
`brokerd` vs. tsdb end index. Repair this by doing timeframe aware index
diffing in `diff_history()` which seems to resolve it. Also, use the
frame-result's `end_dt: datetime` for the loop exit condition.
Sync per-symbol sampler loop start to subscription registers such that
the loop can't start until the consumer's stream subscription is added;
the task-sync uses a `trio.Event`. This patch also drops a ton of
commented cruft.
Further adjustments needed to get parity with prior functionality:
- pass init msg 'symbol_info' field to the `Symbol.broker_info: dict`.
- ensure the `_FeedsBus._subscriptions` table uses the broker specific
(without brokername suffix) as keys for lookup so that the sampler
loop doesn't have to append in the brokername as a suffix.
- ensure the `open_feed_bus()` flumes-table-msg returned sent by
`tractor.Context.started()` uses the `.to_msg()` form of all flume
structs.
- ensure `maybe_open_feed()` uses `tractor.MsgStream.subscribe()` on all
`Flume.stream`s on cache hits using the
`tractor.trionics.gather_contexts()` helper.
Orient shm-flow-arrays around the new idea of a `Flume` which provides
access, mgmt and basic measure of real-time data flow sets (see water
flow management semantics).
- We discard the previous idea of a "init message" which contained all
the shm attachment info and instead send a startup message full of
`Flume.to_msg()`s which are symmetrically loaded on the caller actor
side.
- Create data-flows "entries" for every passed in fqsn such that the consumer gets back
streams and shm for each, now all wrapped in `Flume` types. For now we
allocate `brokermod.stream_quotes()` tasks 1-to-1 for each fqsn
(instead of expecting each backend to do multi-plexing, though we
might want that eventually) as well a `_FeedsBus._subscriber` entry
for each. The pause/resume management loop is adjusted to match.
Previously `Feed`s were allocated 1-to-1 with each fqsn.
- Make `Feed` a `Struct` subtype instead of a `@dataclass` and move all
flow specific attrs to the new `Flume`:
- move `.index_stream()`, `.get_ds_info()` to `Flume`.
- drop `.receive()`: each fqsn entry will now require knowledge of
separate streams by feed users.
- add multi-fqsn tables: `.flumes`, `.streams` which point to the
appropriate per-symbol entries.
- Async load all `Flume`s from all contexts and all quote streams using
`tractor.trionics.gather_contexts()` on the client `open_feed()` side.
- Update feeds test to include streaming 2 symbols on the same (binance)
backend.
This is to prep for multi-symbol feeds and charts so we accept
a sequence of fqsns to the top level entrypoints as well as the
`.data.feed.open_feed()` API (though we're not actually supporting true
multiplexed feeds nor shm lookups per fqsn yet).
Allows starting UI apps and passing the `pikerd` registry socket-addr
args via `--host` or `--port` such that a separate actor tree can be
started by selecting an unused port. This is handy when hacking new
features but while also wishing to run a more stable version of the code
for trading on the same host.
Drop all attempts at rewiring `ViewBox` signals, monkey-patching
relayee handlers, and generally modifying event source public
attributes. Instead take a much simpler approach where the event source
graphics object simply has it's handler dynamically overridden by
a broadcaster function which relays to all consumers using a Python
loop.
The benefits of this much simplified approach include:
- avoiding the tedious and often complex (re)connection of signals between
the source plot and the overlayed consumers.
- requiring zero modification of the public interface of any of the
publisher or consumer `ViewBox`s, no decoration, extra signal
definitions (eg. previous `mouseDragEventRelay` or the like).
- only a single dynamic method override on the event source graphics object
(`ViewBox`) which does the broadcasting work and requires no
modification to handler implementations.
Detailed `.ui._overlay` changes:
- drop `mk_relay_signal()`, `enable_relays()` which removes signal/slot
hacking methodology.
- drop unused `ComposedGridLayout.grid` and `.reverse`, change some
method names: `.insert()` -> `.insert_plotitem()`, `append()` ->
`.append_plotitem()`.
- in `PlotOverlay`, again drop all signal/slot rewiring in
`.add_plotitem()` and instead add our new closure based python-loop in
`broadcast()` routine which is used to override the event-source
object's handler.
- comment out all the auxiliary/want-to-have event source selection
methods for now.
Mainly this involves instantiating our overriden `PlotItem` in a few
places and tweaking type annots. A further detail is that inside
the fsp sub-chart creation code we hide some axes for overlays in the
flows subchart; these were previously somehow hidden implicitly?
Fork out our patch set submitted to upstream in multiple PRs (since they
aren't moving and/or aren't a priority to core) which can be seen in
full from the following diff:
https://github.com/pyqtgraph/pyqtgraph/compare/master...pikers:pyqtgraph:graphics_pin
Move these type extensions into the internal `.ui._pg_overrides` module.
The changes are related to both `pyqtgraph.PlotItem` and `.AxisItem` and
were driven for our need for multi-view overlays (overlaid charts with
optionally synced axis and interaction controls) as documented in the PR
to upstream: https://github.com/pyqtgraph/pyqtgraph/pull/2162
More specifically,
- wrt to `AxisItem` we added lru caching of tick values as per:
https://github.com/pyqtgraph/pyqtgraph/pull/2160.
- wrt to `PlotItem` we adjusted some of the axis management code, namely
adding a standalone `.removeAxis()` and modifying the `.setAxisItems()` logic
to use it in: https://github.com/pyqtgraph/pyqtgraph/pull/2162
as well as some tweaks to `.updateGrid()` to loop through all possible
axes when grid setting.
Details of the original patch to upstream are in:
https://github.com/pyqtgraph/pyqtgraph/pull/2281
Instead of trying to land this we've opted to just copy out that version
of `.debug.Profiler` into our own internals (luckily the class is
entirely self-contained) until such a time when we choose to find
a better dependency as per https://github.com/pikers/piker/issues/337
To make it easier to manually read/decipher long ledger files this adds
`dict` sorting based on record-type-specific (api vs. flex report)
datetime processing prior to ledger file write.
- break up parsers into separate routines for flex and api record
processing.
- add `parse_flex_dt()` for special handling of the weird semicolon
stamps in flex reports.
There never was any underlying db bug, it was a hardcoded timeframe in
the column series write key.. Now we always assert a matching timeframe
in results.
Not only improves startup latency but also avoids a bug where the rt
buffer was being tsdb-history prepended *before* the backfilling of
recent data from the backend was complete resulting in our of order
frames in shm.
Factor the multi-sample-rate region UI connecting into a new helper
`link_views_with_region()` which reads in the shm buffer offsets from
the `Feed` and appropriately connects the fast and slow chart handlers
for the linear region graphics. Add detailed comments writeup for the
inter-sampling transform algebra.
If a history manager raises a `DataUnavailable` just assume the sample
rate isn't supported and that no shm prepends will be done. Further seed
the shm array in such cases as before from the 1m history's last datum.
Also, fix tsdb -> shm back-loading, cancelling tsdb queries when either
no array-data is returned or a frame is delivered which has a start time
no lesser then the least last retrieved. Use strict timeframes for every
`Storage` API call.
Turns out querying for a high freq timeframe (like 1sec) will still
return a lower freq timeframe (like 1Min) SMH, and no idea if it's the
server or the client's fault, so we have to explicitly check the sample
step size and discard lower freq series-results. Do this inside
`Storage.read_ohlcv()` and return an empty `dict` when the wrong time
step is detected from the query result.
Further enforcements,
- both `.load()` and `read_ohlcv()` now require an explicit `timeframe:
int` input to guarantee the time step of the output array.
- drop all calls `.load()` with non-timeframe specific input.
Our default sample periods are 60s (1m) for the history chart and 1s for
the fast chart. This patch adds concurrent loading of both (or more)
different sample period data sets using the existing loading code but
with new support for looping through a passed "timeframe" table which
points to each shm instance.
More detailed adjustments include:
- breaking the "basic" and tsdb loading into 2 new funcs:
`basic_backfill()` and `tsdb_backfill()` the latter of which is run
when the tsdb daemon is discovered.
- adjust the fast shm buffer to offset with one day's worth of 1s so
that only up to a day is backfilled as history in the fast chart.
- adjust bus task starting in `manage_history()` to deliver back the
offset indices for both fast and slow shms and set them on the
`Feed` object as `.izero_hist/rt: int` values:
- allows the chart-UI linked view region handlers to use the offsets
in the view-linking-transform math to index-align the history and
fast chart.
Allows keeping mutex state around data reset requests which (if more
then one are sent) can cause a throttling condition where ib's servers
will get slower and slower to conduct a reconnect. With this you can
have multiple ongoing contract requests without hitting that issue and
we can go back to having a nice 3s timeout on the history queries before
activating the hack.
When a network outage or data feed connection is reset often the
`ib_insync` task will hang until some kind of (internal?) timeout takes
place or, in some (worst) cases it never re-establishes (the event
stream) and thus the backend needs to restart or the live feed will
never resume..
In order to avoid this issue once and for all this patch implements an
additional (extremely simple) task that is started with the real-time
feed and simply waits for any market data reset events; when detected
restarts the `open_aio_quote_stream()` call in a loop using
a surrounding cancel scope.
Been meaning to implement this for ages and it's finally working!
Allows for easier restarts of certain `trio` side tasks without killing
the `asyncio`-side clients; support via flag.
Also fix a bug in `Client.bars()`: we need to return the duration on the
empty bars case..
This allows the history manager to know the decrement size for
`end_dt: datetime` on the next query if a no-data / gap case was
encountered; subtract this in `get_bars()` in such cases. Define the
expected `pendulum.Duration`s in the `.api._samplings` table.
Also add a bit of query latency profiling that we may use later to more
dynamically determine timeout driven data feed resets. Factor the `162`
error cases into a common exception handler block.
When we get a timeout or a `NoData` condition still return a tuple of
empty sequences instead of `None` from `Client.bars()`. Move the
sampling period-duration table to module level.
It doesn't seem to be any slower on our least throttled backend
(binance) and it removes a bunch of hard to get correct frame
re-ordering logic that i'm not sure really ever fully worked XD
Commented some issues we still need to resolve as well.
Manual tinker-testing demonstrated that triggering data resets
completely independent of the frame request gets more throughput and
further, that repeated requests (for the same frame after cancelling on
the `trio`-side) can yield duplicate frame responses. Re-work the
dual-task structure to instead have one task wait indefinitely on the
frame response (and thus not trigger duplicate frames) and the 2nd data
reset task poll for the first task to complete in a poll loop which
terminates when the frame arrives via an event.
Dirty deatz:
- make `get_bars()` take an optional timeout (which will eventually be
dynamically passed from the history mgmt machinery) and move request
logic inside a new `query()` closure meant to be spawned in a task
which sets an event on frame arrival, add data reset poll loop in the
main/parent task, deliver result on nursery completion.
- handle frame request cancelled event case without crash.
- on no-frame result (due to real history gap) hack in a 1 day decrement
case which we need to eventually allow the caller to control likely
based on measured frame rx latency.
- make `wait_on_data_reset()` a predicate without output indicating
reset success as well as `trio.Nursery.start()` compat so that it can
be started in a new task with the started values yielded being
a cancel scope and completion event.
- drop the legacy `backfill_bars()`, not longer used.
Adjust all history query machinery to pass a `timeframe: int` in seconds
and set default of 60 (aka 1m) such that history views from here forward
will be 1m sampled OHLCV. Further when the tsdb is detected as up load
a full 10 years of data if possible on the 1m - backends will eventually
get a config section (`brokers.toml`) that allow user's to tune this.
The `Store.load()`, `.read_ohlcv()` and `.write_ohlcv()` and
`.delete_ts()` now can take a `timeframe: Optional[float]` param which
is used to look up the appropriate sampling period table-key from
`marketstore`.
Allow data feed sub-system to specify the timeframe (aka OHLC sample
period) to the `open_history_client()` delivered history fetching API.
Factor the data keycombo hack into a new routine to be used also from
the history backfiller code when request latency increases; there is
a first draft at trying to use the feed reset to speed up 1m frame
throttling by timing out on the history frame response, but it needs
a lot of fine tuning.
This is a simpler (and oddly more `trio`-nic and/or SC) way to handle
the cancelled-before-acked race for order dialogs. Will allow keeping
the `.req` field as solely an `Order` msg.
When the client is faster then a `brokerd` at submitting and cancelling
an order we run into the case where we need to specify that the EMS
cancels the order-flow as soon as the brokerd's ack arrives. Previously
we were stashing a `BrokerdCancel` msg as the `Status.req` msg (to be
both tested for as a "already cancelled" and sent immediately on ack arrival to
the broker), but for such
cases we can't use that msg to find the fqsn (since only the client side
msgs have it defined) which is required by the new
`Router.client_broadcast()`.
So, Since `Status.req` is supposed to be a client-side flow msg anyway,
and we need the fqsn for client broadcasting, we change this `.req`
value to the client's submitted `Cancel` msg (thus rectifying the
missing `Router.client_broadcast()` fqsn input issue) and build the
`BrokerdCancel` request from that `Cancel` inline in the relay loop
from the `.req: Cancel` status msg lookup.
Further we allow `Cancel` msgs to define an `.account` and adjust the
order mode loop to expect `Cancel` source requests in cancelled status
updates.
Except for paper accounts (in which case we need a trades dialog and
paper engine per symbol to enable simulated clearing) we can rely on the
instrument feed (symbol name) to be the caching key. Utilize
`tractor.trionics.maybe_open_context()` and the new key-as-callable
support in the paper case to ensure we have separate paper clearing
loops per symbol.
Requires https://github.com/goodboy/tractor/pull/329
With the refactor of the dark loop into a daemon task already-open order
relaying from a `brokerd` was broken since no subscribed clients were
registered prior to the relay loop sending status msgs for such existing
live orders. Repair that by adding one more synchronization phase to the
`Router.open_trade_relays()` task: deliver a `client_ready: trio.Event`
which is set by the client task once the client stream has been
established and don't start the `brokerd` order dialog relay loop until
this event is ready.
Further implementation deats:
- factor the `brokerd` relay caching back into it's own `@acm` method:
`maybe_open_brokerd_dialog()` since we do want (but only this) stream
singleton-cached per broker backend.
- spawn all relay tasks on every entry for the moment until we figure
out what we're caching against (any client pre-existing right, which
would mean there's an entry in the `.subscribers` table?)
- rename `_DarkBook` -> `DarkBook` and `DarkBook.orders` -> `.triggers`
This enables "headless" dark order matching and clearing where an `emsd`
daemon subactor can be left running with active dark (or other
algorithmic) orders which will still trigger despite to attached-controlling
ems-client.
Impl details:
- rename/add `Router.maybe_open_trade_relays()` which now does all work
of starting up ems-side long living clearing and relay tasks and the
associated data feed; make is a `Nursery.start()`-able task instead of
an `@acm`.
- drop `open_brokerd_trades_dialog()` and move/factor contents into the
above method.
- add support for a `router.client_broadcast('all', msg)` to wholesale
fan out a msg to all clients.
Establishes a more formalized subscription based fan out pattern to ems
clients who subscribe for order flow for a particular symbol (the fqsn
is the default subscription key for now).
Make `Router.client_broadcast()` take a `sub_key: str` value which
determines the set of clients to forward a message to and drop all such
manually defined broadcast loops from task (func) code. Also add
`.get_subs()` which (hackily) allows getting the set of clients for
a given sub key where any stream that is detected as "closed" is
discarded in the output. Further we simplify to `Router.dialogs:
defaultdict[str, set[tractor.MsgStream]]` and `.subscriptions` as maps
to sets of streams for much easier broadcast management/logic using set
operations inside `.client_broadcast()`.
This patch was originally to fix a bug where new clients who
re-connected to an `emsd` that was running a paper engine were not
getting updates from new fills and/or cancels. It turns out the solution
is more general: now, any client that creates a order dialog will be
subscribing to receive updates on the order flow set mapped for that
symbol/instrument as long as the client has registered for that
particular fqsn with the EMS. This means re-connecting clients as well
as "monitoring" clients can see the same orders, alerts, fills and
clears.
Impl details:
- change all var names spelled as `dialogues` -> `dialogs` to be
murican.
- make `Router.dialogs: dict[str, defaultdict[str, list]]` so that each
dialog id (oid) maps to a set of potential subscribing ems clients.
- add `Router.fqsn2dialogs: dict[str, list[str]]` a map of fqsn entries to
sets of oids.
- adjust all core task code to make appropriate lookups into these 2 new
tables instead of being handed specific client streams as input.
- start the `translate_and_relay_brokerd_events` task as a daemon task
that lives with the particular `TradesRelay` such that dialogs cleared
while no client is connected are still processed.
- rename `TradesRelay.brokerd_dialogue` -> `.brokerd_stream`
- broadcast all status msgs to all subscribed clients in the relay loop.
- always de-reg each client stream from the `Router.dialogs` table on close.
Not sure what exactly happened but it seemed clears weren't working in
some cases without this, also there's no point in spinning the simulated
clearing loop if we're handling a non-clearing tick type.
We haven't been using it for a while and the supposed (remembered)
latency issue on interaction doesn't seem existing after applying the
cache mode. This allows dropping some internal state-logic and generally
simplifying the show-on-hover checks.
Further add `.show_markers()` and `.hide_markers()` as explicit methods
that can be called externally by UI business logic.
Bit of a face palm but obviously `LevelLine.delete()` also removes any
`._marker` from the view which makes it disappear permanently when
moving from non-zero to zero to non-zero positions.. We don't really
need to delete the line since it can be re-used so just remove that
code.
Further this patch removes marker style setting logic from within the
`pp_line()` factory and instead expects the caller to set the correct
"direction" (for long / short) afterward.
- Every time a symbol is switched on chart we need to wait until the
search bar sidepane has been added beside the slow chart before
determining the offset for the pp line's arrow/labels; trigger this in
`GodWidget.load_symbol()` -> required monkeypatching on a
`.mode: OrderMode` to the `.rt_linked` for now..
- Drop the search pane widget removal from the current linked chart,
seems faster?
- On the slow chart override the `LevelMarker.scene_x()` callback to
adjust for the case where no L1 labels are shown beside the y-axis.
Also adds a `GodWidget.resize_all()` helper method which resizes all
sub-widgets and charts to their default ratios and/or parent-widget
dependent defaults using the detected available space on screen. This is
a "default layout" config method that eventually we'll probably want
allow users to customize.
In other words instead of some static view size previously determined by
the accompanying (slow) chart's height, (recursively) calculate the
number of displayed rows and compute the minimal height needed. This
still caps the view at the height of the chart such that the view will
switch to scroll bar mode when too many results are shown and can't all
be fit in the vertical space.
Deats:
- add a ``CompleterView.iter_df_rows()`` which recursively iterates all
rows in depth-first order making it simple to compute the absolute
number of result rows in view and thus the minimal number of pixels to
show all results.
- always pass the height in the `.on_resize()` handler to ensure
triggering the height logic when new results are generated in the
search loop.
Scales the "view" instance that holds search results to the size of the
accompanying "slow chart" for which the search pane is a "sidepane".
A lot of mucking about was required due to resizing of the view
seemingly feeding back into window resizing and further implementing the
sizing logic such that the parent `QSplitter` can be resized as the
user's whim as well.
Details,
- add a `CompleterView._init: bool` which is set once (and only once)
after startup where the first display of the current symbol/feed is
shown allowing and a single *width* padding applied once at startup
to ensure we don't have an awkward line to the right of the longest
result.
- in `.resize_to_results()` only apply a minimum height to the view
using `.setMinimumHeight()` with a down-scaled (`0.91` for now) height
value from input.
- re-implement `CompleterView.show_matches()` to accept and optional
width, heigh tuple and when not supplied pull the slow chart's
dimensions and pass as input to the resize method.
- Make `SearchWidget` x dim sizing policy "fixed".
- register the `SearchWidget` for resize events with god.
- add `.show_only_cache_entries()` for easy results clearing.
- add `.space_dims()` to retrieve slow linked-charts dimensions.
- implement `SearchWidget.on_resize()` which is the caller of all the
previously mentioned resizing routines.
- do resizing and cache entry showing on search loop startup and be sure
to clear to cache when the user selects a symbol-feed with Enter.
It ended up being what'd you expect, races on the accessing shm buffer
data by the UI during the whole "mega-async-startup-everything" phase XD
So we add the following list of ad-hoc startup steps:
- do `.default_view()` on the slow chart after the fast chart is mostly
fully spawned with the intention being to capture the state where the
historical buffer is mostly loaded before sizing the view to the
graphical form of the data.
- resize slow chart sidepanes from the fast chart just before sleeping
forever (and after order mode has booted).
Turns out god widget resizes aren't triggered implicitly by window
resizes, so instead, hook into the window by moving what was our useless
method to that class. Further we explicitly define and declare that our
window has a `.godwidget: GodWidget` and set it up in the bootstrap
phase - in `run_qutractor()` during `trio` guest mode configuration.
Further deatz:
- retype the runtime/bootstrap routines to take a qwidget "type" not an
instance, and drop the whole implicit `.main_widget` stuff.
- delegate into the `GodWidget.on_win_resize()` for any window resize
which then triggers all the custom resize callbacks we already had in
place.
- privatize `ChartnPane.sidepane` so that it can't be mutated willy
nilly without calling `.set_sidepane()`.
- always adjust splitter sizes inside `LinkeSplits.add_plot()`.
More or less moves all the UI related position "nav" logic and graphics
item management into a new `._position.Nav` composite type + api for
high level mgmt of position graphics indicators across multiple charts
(fast and slow).
The slow (history) chart requires it's own y-range checker logic which
needs to be run in 2 cases:
- the last datum is in view and goes outside the previous mx/mn in view
- the chart is incremented a step
Since we need this duplicate logic this patch also factors the incremental
graphics update info "reading" into a new `DisplayState.incr_info()`
method that can be configured to a chart and input state and returns all
relevant "graphics update measure" in a tuple (for now).
Use this method throughout the rest of the display loop for both fast
and slow chart checks and in the `increment_history_view()` slow chart
task.
Use the new `Feed.get_ds_info()` method in a poll loop to definitively
get the inter-chart sampling info and avoid races with shm buffer
backfilling.
Also, factor the history increment closure-task into
`graphics_update_loop()` which will make it clearer how to factor
all the "should we update" logic into some `DisplayState` API.
If you spawn a brokerd set and no `ib` data feed was started (via our
`.data.feed.Feed` api) then there will be no active client loaded and
thus wont' be connected. So in these cases just return nothing, and
I guess we'll figure out real connection failures later?
Add an update call to the display loop to consistently update the last
datum in the history view chart. Compute the inter-chart sampling ratio
and use it to sync the linear region.
Add a first draft of a working `pyqtgraph.LinearRegionItem` link between
a history view chart (+ data set) and the normal real-time "HFT" chart
set.
Add the history view (aka more downsampled data view) chart set to the
rt/hft set's splitter as it's "first widget". Hook up linear region
callbacks to enable syncing between charts including compenstating for
the downsampling rate ration (in this case hardcoded 60 since 1s to 1M,
but we'll actually compute it going forward obvs).
More to come dawgys..
Adds an additional `GodWidget.hist_linked: LinkedSplits` alongside the
renamed `.rt_linked` to enable 2 sets of linked charts with different
sampled data sets/flows. The history set is added without "all the
fixins" for now (i.e. no order mode sidepane or search integration) such
that it is merely a top level chart which shows a much longer term
history and can be added to the UI via embedding the entire history
linked-splits instance into the real-time linked set's splitter.
Further impl deats:
- adjust the `GodWidget._chart_cache: dict[str, tuple]]` to store both
linked split chart sets per symbol so that symbol switching will
continue to work with the added history chart (set).
- rework `.load_symbol()` to operate on both the real-time (HFT) chart
set and the history set.
- rework `LinkedSplits.set_split_sizes()` to compensate for the history
chart and do more detailed height calcs arithmetic to make it appear
by default as a minor sub-chart.
- adjust `LinkedSplits.add_plot()` and `ChartPlotWidget` internals to allow
adding a plot without a sidepane and/or container `ChartnPane`
composite widget by checking for a `sidepane == False` input.
- make `.default_view()` accept a manual y-axis offset kwarg.
- adjust search mode to provide history linked splits to
`.set_chart_symbol()` call.
As part of supporting a "history view" chart which shows downsampled
datums alongside our 1s (or higher) sampled OHLC we need a separate
buffer to store a the slower history from broker backends. This begins
that design by allocating 2 buffers:
- `rt_shm: ShmArray` which maps to a `/dev/shm/` file with `_rt` suffix
- `hist_shm: ShmArray` which maps to a file with `_hist` suffix
Deliver both of these shms back from both `manage_history()` and load
them as `Feed.rt_shm`/`.hist_shm` on the client side.
Impl deats:
- init the rt buffer with the first datum from loaded history and
assign all OHLC values to that row's 'close' and the vlm to 0.
- pass the hist buffer to the backfiller task
- only spawn **one** global sampler array-row increment task per
`brokerd` and pass in the 1s delay which we presume is our lowest
OHLC sample rate for now.
- drop `open_sample_step_stream()` and just move its body contents into
`Feed.index_stream()`
Instead of worrying about the increment period per shm subscription,
just use the value passed as input and presume the caller knows that
only one task is necessary and that the wakeup (sampling) period should
be the shortest that is needed.
It's very unlikely we don't want at least a 1s sampling (both in terms
of task switching cost and general usage) which will eventually ship as
the default "real-time" feed "timeframe". Further, this "fast" increment
sampling task can handle all lower sampling periods (eg. 1m, 5m, 1H)
based on the current implementation just the same.
Also, add a global default sample period as `_defaul_delay_s` for use in
other internal modules.
Clearly, the linter didn't help us here.. but, just pass the
`brokerd` time for now in the `.broker_time` field; we can't get it from
the fill-case incremental updates in the `openOrders` sub. Add some
notes about this and how we might approach for backends with this
limitation.
This fixes a regression added after moving the msg parsing to later in
the order mode startup sequence. The `Allocator` needs to be configured
*to* the initial pos otherwise default settings will show in the UI..
Move the startup config logic from inside `mk_allocator()` to
`PositionTracker.update_from_pp()` and add a flag to allow setting the
`.startup_pp` from the current live one as is needed during initial
load.
In the short case (-ve size) we had a bug where the last sub-slots worth
of exit size would never be limited to zero once the allocator limit pos
size was hit (i.e. you could keep going more -ve on the pos,
exponentially per slot over the limit). It's a simple fix, just
a `max()` around the `l_sub_pp` var used in the next-step-size calc.
Resolves#392
Turns out we were putting too many brokername suffixes in the symbol
field and thus the order mode msg parser wasn't matching the current
asset to said msgs correctly and pps weren't being shown...
This repairs that plus simplifies the order mode initial pos msg loading
to just delegate into `process_trade_msg()` just as is done for
real-time msg updates.
If a setting fails to apply try to log an error msg and revert to the
previous setting by not applying the UI read-update until after the new
`SettingsPane.apply_setting()` call. This prevents crashes when the user
tries to give bad inputs on editable allocator fields.
Previously we only simulated paper engine fills when the data feed
provide L1 queue-levels matched an execution. This patch add further
support for clear-level matches when there are real live clears on the
data feed that are faster/not synced with the L1 (aka usually during
periods of HFT).
The solution was to simply iterate the interleaved paper book entries on
both sides for said tick types and instead yield side-specific predicate
per entry.
Not entirely sure why this all of a sudden became a problem but it seems
price changes on order edits were sometimes resulting in key errors when
modifying paper book entries quickly. This changes the implementation to
not care about matching the last price when keying/popping old orders
and use `bidict`s to more easily pop cleared orders in the paper loop.
When the paper engine is used it seems we can definitely hit races where
order ack msgs arrive close enough to status messages that `trio`
schedules the status processing before the acks. In such cases we want
to be tolerant and not crash but instead warn that we got an
unknown/out-of-order msg.
Tyler Goodlet
goodboy
Guillermo Rodriguez
Esmeralda Gallardo
algorandpa
algorandpa
Guillermo Rodriguez