timeBlend {openblender} | R Documentation |
Get Dataframe with DataToBlend ordered time by sending a list or Series of Anchor timestamps. Consult documentation here.
timeBlend( token, anchor_ts, blend_source, blend_type = "closest_observation", direction = "time_prior", interval_output = "count", ts_restriction = NULL, oblender = FALSE, interval_size = 3600, consumption_confirmation = "off", missing_values = "raw", data_format = "dataframe" )
token |
Obtained in openblender.io |
anchor_ts |
Array of Unix timestamp values to blend data to. |
blend_source |
JSON object of Data To Blend configuration obtained from a data source in OpenBlender.io |
blend_type |
Type of blend to gather data: 'closest_observation', 'agg_in_intervals', (Default: closest_observation). |
direction |
(Only if blend_type = 'agg_in_intervals'). Select chronological direction of observation or interval to aggregate from 'blend source' to 'anchor': 'time_prior': Gather data from blend source prior or at the same time as each observation from anchor. 'time_prior_strict': Gather data from blend source strictly prior to each observation from anchor. 'time_post': Gather data subsequent to or at the same time than each observation from anchor. 'time_post_strict': Gather data strictly subsequent to each observation from anchor. 'time_prior_and_post': Gather data prior and subsequent to each observation from anchor. Default (time_prior) |
interval_output |
(Only if blend_type = 'agg_in_range'), options: 'count' (any type of feature). 'sum' 'avg' (for numerical features). 'text', 'text_list' (for text features). (Default = count) |
ts_restriction |
None |
oblender |
Omit this parameter |
interval_size |
(Only if blend_type = 'agg_in_intervals'). Time interval size in seconds to be used with the selected agg_interval_order. |
consumption_confirmation |
"on" or "off", default "on" |
missing_values |
Treatment of missing values. 'raw': Ignore missing values (return NaN). 'impute': Impute missing values. (Default 'raw') |
data_format |
Default "Dataframe" |
A Blended dataset intersected with your anchor in time, success/error message when you insert observations or the list of observations requested.
To see more details go to openblender.io
## Not run: #Blend a dataset #You provide df_anchor data blend_source <- list( id_dataset="5f063dda198b5854284dae58", feature="price", dataset_filter=c(list(low=list("$gt"=300))) ) df_blend <- openblender::timeBlend(token = "YOUR_TOKEN", anchor_ts = df_anchor$timestamp, blend_type = "agg_in_intervals", direction = "time_prior", interval_output = "avg", interval_size = 60 * 60 * 24 * 1, blend_source = blend_source, missing_values = "impute") ## End(Not run)