DataViews - Functions

DVVertical(col::Collection{A}, dv, args...) where A <: AssetsDisc

Create a vertical dataview that splits timeseries data into "before" and "after" segments relative to each time point.

Arguments

• col::Collection{A}: A collection of assets with discretized samples
• dv: A function that creates a dataview for each asset
• args...: Additional arguments passed to the dataview function

Returns

A DVVertical struct containing split views for each asset and date in the collection.

Access Patterns

For a returned dvv::DVVertical:

• dvv[ticker][date][i] - Tuple of (before, after) views split at i-th timepoint
• dvv[ticker][date][1][end] - Value at evaluation time
• dvv[ticker][date][1][end-1] - Value one tick before evaluation
• dvv[ticker][date][2][1] - Value one tick after evaluation

DVCheckBackAvailable(tsdl::TSDLink, field::Symbol)

Returns the maximum number of time steps that can be gone back, which is one less than the start index of the time series data.

DVLayerDictKey(asset::Asset, dict::Symbol, key::Union{Int64, Symbol},

layertype::Type{T}=AssetLayer) where T <: AssetLayer View of a key in a field dict of type Dict for layer of type `layertypeinassetThe datetimes are matched to those in theassetso thatDVLayerDictKey[x]gives the value in the dict of the layer field value corresponding toasset[x]`

DVLayerField(asset::Asset, field::Symbol, layer_type::Type{T}=AssetLayer) where T <: AssetLayer

View of a field field for layer of type layer_type in asset The datetimes are matched to those in the asset so that DVLayerField[x] gives the layer field value corresponding to asset[x]

DVLayerGetFunc(asset::Asset, func::Function, layer_type::Type{T}=AssetLayer) where T <: AssetLayer

The function func applied to layer of type layer_type in asset The datetimes are matched to those in the asset so that DVLayerGetFunc[x] gives the function func func(asset.some_field[x])

DVLayerGetFuncPar(asset::Asset, func::Function, par::Any, layer_type::Type{T}=AssetLayer) where T <: AssetLayer

The function func that takes the additional parameter par applied to layer of type layer_type in asset The datetimes are matched to those in the asset so that DVLayerGetFuncPar[x] gives func(asset.some_field[x], par)

DVNonRepaintingLayerDictKey(asset::Asset, dict::Symbol, key::Union{Int64, Symbol}, layer_type::Type{T}=AssetLayer) where T <: AssetLayer

View of a key in a field dict of type Dict for layer of type layer_type in asset

The datetimes are matched to those in the asset in a non repainting manner so that DVNonRepaintingLayerDictKey[1:end] equals the value in the dict of the layer field assets from minimum to maximum of all datetimes inasset`. Useful for constructing non repainting layers.

DVNonRepaintingTimeSeries(asset::Asset, field::Symbol)

View of data from associated time series in SFDB with the name symbol.

The datetimes are matched to those in the asset in a non repainting manner so that DVNonRepaintingTimeSeries[1:end] equals the time series assets from minimum to maximum of all datetimes inasset`. Useful for constructing non repainting layers.

DVNonRepaintingTimeSeriesBack(asset::Asset, field::Symbol, back::Int64)

View of data from associated time series in SFDB with the name symbol.

The datetimes are matched to those in the asset in a non repainting manner so that DVNonRepaintingTimeSeriesBack[1:end] equals the time series assets from minimum to maximum of all datetimes inasset. Useful for constructing non repainting layers. Theback` variable represents the number of time steps to go back in time from the earlest datetime in the asset.

DVPriceSeries(asset::Asset, field::Symbol)

View of data from associated price series in SFDB.

The datetimes are matched to those in the asset so that DVTimeSeries[x][y] gives the time series value corresponding to asset[x][y]

DVTimeSeries(asset::Asset, field::Symbol)

View of data from associated time series in SFDB with the name field.

The datetimes are matched to those in the asset so that DVTimeSeries[x][y] gives the time series value corresponding to asset[x][y]

FutureAdjClose(asset::Asset)

A view of adjusted close prices from the evaluation datetime and forward. The evaluated datetime is included. FutureAdjClose[x] gives a view of future adjusted close values from the evaluate datetime of asset[x]

FutureAdjCloseDay(asset::Asset)

A view of adjusted close prices from the evaluation date and forward at a daily resolution. The evaluated date is included. FutureAdjCloseDay[x] gives a view of future adjusted close values from the evaluate date of asset[x]. Useful for calculations of daily returns from decisions made on higher resolution time frames.

FutureAdjCloseEvalToDay(asset::Asset)

A view of adjusted close prices from the evaluation datetime and then forward at a daily resolution. FutureAdjCloseDayToEval[x] gives a view of future adjusted close values from the evaluate datetime of asset[x] where FutureAdjCloseEvalToDay[x][1] is the adj close of the evaluated datetime and FutureAdjCloseEvalToDay[2] is the adj close of the associated date. Then follows adj close on a daily resolution. Useful for calculations of daily returns from decisions made on higher resolution time frames.