Package 'vmxr'

Description

A sentinel for an explicit "All specimens" selection. Pass it as pk_specimen to vmx_nca(), vmx_modeling_options() or vmx_model_build() to pool PK observations across every specimen, or as the specimen of vmx_pd_marker() to pool one PD marker across its specimens.

Usage

vmx_all_specimens()

Details

All specimens is never a default; a call that needs a specimen and gets none fails rather than choosing one.

Value

A vmx_all_specimens object.

See Also

vmx_pd_marker()

Examples

vmx_all_specimens()

Description

GET /studies/{std_id}/analysis-log — the complete unified newest-first event feed, with a kind discriminator per row.

Usage

vmx_analysis_log(
  study,
  kind = NULL,
  event_type = NULL,
  outcome = NULL,
  severity = NULL,
  since = NULL,
  resource = NULL,
  client = vmx_client(),
  event_code = NULL,
  requires_staff_review = NULL
)

Arguments

Value

A tibble containing all matching events, newest first.

Description

Resolves connection config with the precedence: explicit args -> VMX_API_BASE_URL / VMX_API_TOKEN environment variables -> ⁠~/.Renviron⁠ -> a classed error. The token is stored on the object but redacted in print() and never logged.

Usage

vmx_client(base_url = NULL, token = NULL, ...)

Arguments

Details

When no token is supplied (neither ⁠token=⁠ nor VMX_API_TOKEN), the client auto-authenticates via OIDC: it reads the token cached by vmx_login() (⁠~/.config/vmx/oidc-token.json⁠), refreshing it silently when expired, and – in an interactive session – running vmx_login() if there is no usable cached token. This requires VMX_OIDC_ISSUER + VMX_OIDC_CLIENT_ID to be set; otherwise a vmx_auth_error names both auth methods.

The bearer token is resolved per request, not frozen at construction: a PAT is constant, but an OIDC access token is short-lived (~5-10 min), so a long-lived con <- vmx_client() re-resolves each request and silently refreshes from the cached refresh token when the access token is near expiry. This lets a reused client keep working across a full analysis session without re-login (GEN-2344).

Value

An object of class vmx_client.

Description

Fetch one data version

Usage

vmx_data_version(id, client = vmx_client())

Arguments

Value

A vmx_data_version.

Description

Archive a data version

Usage

vmx_data_version_archive(dv, reason = NULL, client = vmx_client())

Arguments

Value

The updated vmx_data_version.

Description

Starts a format job over an explicit upload composition (POST /datasets/{ds_id}/data-versions). Returns the in-flight prep-status; poll it with vmx_wait().

Usage

vmx_data_version_create(
  dataset,
  uploads,
  prior_config = NULL,
  client = vmx_client()
)

Arguments

Value

A vmx_prep_status for the new job.

Description

Fetches the export envelope (GET /data-versions/{id}/export). Pass dest to obtain the bundle: it is streamed to that path via the server's signed download URL. That URL is short-lived — its life is measured in seconds — and is not returned. Without dest, the parsed envelope is returned with its download_url removed, leaving the export's details (its size, its file list, its data version).

Usage

vmx_data_version_export(dv, dest = NULL, client = vmx_client())

Arguments

Value

Without dest, the export envelope (list) minus its short-lived download_url. With dest, dest invisibly once the bundle is streamed.

Description

GET /data-versions/{id}/tables/{domain} — returns the formatter's prepared domain table as a tibble (columns typed per the server's column metadata, which is attached as the "columns" attribute). gen_subject_uuid is the canonical subject join key.

Usage

vmx_data_version_table(
  dv,
  domain = c("subjects", "pk", "dosing", "pd", "labs", "covariates"),
  time_basis = NULL,
  client = vmx_client()
)

Arguments

Details

API 0.3 projects every analytical table on a time basis and requires the time_basis query parameter. When time_basis is NULL, the DataVersion's recommended basis is used (fetching the DataVersion first if only an id was given). A v0.3 DataVersion with no recommended basis (no basis has PK-eligible subjects) needs an explicit time_basis; only a DataVersion without a time_bases map (an API 0.2 server) is fetched without the parameter.

Value

A tibble. The selected basis is attached as the "time_basis" attribute when one was requested or echoed.

Description

Unarchive a data version

Usage

vmx_data_version_unarchive(dv, client = vmx_client())

Arguments

Value

The updated vmx_data_version.

Description

The modeling-eligibility filter maps onto whichever query param the served API contract accepts. The 0.2.2 API filters on a single basis-agnostic eligible_for_modeling boolean; the 0.3 API replaced it with the basis-scoped pk_eligible_for_modeling_after_qc filter, which requires a companion time_basis (api-contract §5.4). Because a 0.3 server silently drops the unknown 0.2.2 param and returns the full unfiltered list, the served shape is detected first — from /health's api_contract_version (§5.13) — and the filter is mapped accordingly; a 0.3 server asked to filter without a time_basis errors rather than returning an unfiltered list.

Usage

vmx_data_versions(
  treatment = NULL,
  study = NULL,
  include_archived = FALSE,
  eligible_for_modeling = NULL,
  client = vmx_client(),
  time_basis = NULL
)

Arguments

Value

A tibble containing all matching data versions.

Description

Fetch one dataset

Usage

vmx_dataset(id, client = vmx_client())

Arguments

Value

A vmx_dataset.

Description

POST /datasets/{ds_id}/cancel — terminates the format job and mints a cancelled DataVersion.

Usage

vmx_dataset_cancel(dataset, client = vmx_client())

Arguments

Value

The updated vmx_prep_status.

Description

Not implemented: the API's dataset-files listing does not expose per-file download URLs. Use vmx_data_version_export() to pull a curated bundle.

Usage

vmx_dataset_download(dataset, dest = ".", client = vmx_client())

Arguments

Value

Not yet implemented.

Description

List the files in a dataset

Usage

vmx_dataset_files(dataset, client = vmx_client())

Arguments

Value

A tibble containing all files in the dataset.

Description

Returns the dataset's allowlisted tag map as a two-column (key, value) tibble. Reads the tags off a vmx_dataset object when given one, else fetches the dataset.

Usage

vmx_dataset_tags(dataset, client = vmx_client())

Arguments

Value

A tibble with key and value columns.

Description

List datasets

Usage

vmx_datasets(study = NULL, treatment = NULL, client = vmx_client())

Arguments

Value

A tibble containing all matching datasets.

Description

Dosing events table

Usage

vmx_dosing(dv, time_basis = NULL, client = vmx_client())

Arguments

Value

A tibble.

Description

POST /model-fits/{mf_id}/simulation-dosing-inputs.

Usage

vmx_dosing_input(
  fit,
  dosing_text,
  scenario_name,
  client = vmx_client(),
  wait = FALSE,
  ...
)

Arguments

Value

A vmx_dosing_input (carries dosing_input_id).

Description

Dosing-input status

Usage

vmx_dosing_input_status(dosing_input, client = vmx_client())

Arguments

Value

A vmx_dosing_input.

Description

One row per estimate, preserving the server-selected point statistic, interval kind, interval level, and tagged estimate metadata.

Usage

vmx_fit_global_estimates(fit, client = vmx_client())

Arguments

Value

A tibble.

Description

Reshapes the PK block into one row per observation. The returned prediction columns retain the server-selected point statistic and interval semantics. A named list of equivalently reshaped PD-marker tibbles is attached as the "pd_markers" attribute; units and marker references remain attributes on their respective tibbles.

Usage

vmx_fit_obs_vs_pred(fit, client = vmx_client())

Arguments

Value

A tibble (one row per PK observation).

Description

One row per subject x estimate. value_statistic, interval_kind, and interval_level preserve the server-selected estimate semantics; interval_lower and interval_upper are the corresponding bounds. Tagged estimate metadata is retained as columns rather than interpreted by vmxr.

Usage

vmx_fit_subject_estimates(fit, client = vmx_client())

Arguments

Value

A tibble.

Description

Returns the parsed Visual Predictive Check artifact. Its subject and dose-group channels contain model-implied response trajectories with the server-provided point statistic and interval. The nested wire shape is retained verbatim.

Usage

vmx_fit_vpc(fit, client = vmx_client())

Arguments

Value

A list (the parsed artifact).

Description

Calls GET /health.

Usage

vmx_health(client = vmx_client())

Arguments

Value

Invisibly, a vmx_health object (status, version, api_contract_version, engine, engine_version); raises a classed error if the API is unreachable or unhealthy.

Description

Runs the OIDC device-code flow (RFC 8628, via httr2::oauth_flow_device()) against the workspace's Authentik provider and caches the resulting token so later vmx_client() calls authenticate automatically. Endpoints are resolved from the issuer's .well-known/openid-configuration discovery document; the flow uses a public client (no secret, no PKCE) and requests offline_access so Authentik issues a refresh token.

Usage

vmx_login(issuer = NULL, client_id = NULL, scopes = NULL, cache_path = NULL)

Arguments

Details

The token is written as plain JSON to ⁠~/.config/vmx/oidc-token.json⁠ (the same path and shape the vmx CLI uses, with 0600 permissions), so one vmx_login() serves both R and the terminal CLI and the session survives a fresh R process or workspace pod restart. Because the refresh token is persisted on the home PVC, you log in once per provider-configured refresh-token lifetime.

Configuration is read from environment variables (matching the CLI): VMX_OIDC_ISSUER, VMX_OIDC_CLIENT_ID, and optionally VMX_OIDC_SCOPES. Workspace deployments provision the issuer and client id; vmxr does not assume values from a different workspace or environment.

Value

Invisibly, the cached token (a list; the access/refresh tokens are secret and never printed).

Description

Reads the selections the API accepted for a resource (⁠GET .../marker-selection⁠). These are what the analysis actually used and are never recomputed from the data version's current metadata.

Usage

vmx_nca_marker_selection(nca, client = vmx_client())

vmx_model_build_marker_selection(run, client = vmx_client())

vmx_model_fit_marker_selection(fit, client = vmx_client())

Arguments

Details

• vmx_nca_marker_selection() returns list(pk_specimen).

• vmx_model_build_marker_selection() returns list(pk_specimen, pd_markers), where each PD entry carries gen_uuid, either specimen_type or all_specimens = TRUE, and direction. A run accepted before PD specimens were selectable may report only gen_uuid and direction.

• vmx_model_fit_marker_selection() returns list(pk_specimen, pd_marker); pd_marker is NULL for a PK fit, and for a fit saved before specimen selection it may carry only gen_uuid.

pk_specimen is list(specimen_type = "...") or list(all_specimens = TRUE). An analysis accepted before PK specimens were selectable reports all specimens.

Value

A named list, as described above.

Description

pk_specimen is required: one PK specimen type (for example "plasma") or vmx_all_specimens(). vmxr never fills it in, not even from the data version's recommendation; a call without it fails and lists the specimens available. Each PD marker is a vmx_pd_marker() naming one specimen or all specimens, with its direction. Two specimens of one marker may be selected as separate entries. Read the accepted selections back with vmx_model_build_marker_selection().

Usage

vmx_model_build(
  data_version,
  time_basis,
  pk_specimen,
  pd_marker = NULL,
  covariate = NULL,
  idempotency_key = NULL,
  retried_from = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_model_build_run.

Description

Download build-run artifacts

Usage

vmx_model_build_artifacts(run, dest = ".", client = vmx_client())

Arguments

Value

Not yet implemented.

Description

Cancel a build run

Usage

vmx_model_build_cancel(run, client = vmx_client())

Arguments

Value

A vmx_model_build_run.

Description

Build-run events (SSE stream)

Usage

vmx_model_build_events(run, client = vmx_client())

Arguments

Value

Not yet implemented.

Description

Markdown export of a build run

Usage

vmx_model_build_export(run, client = vmx_client())

Arguments

Value

The export markdown as a length-1 character vector.

Description

Build-run logs

Usage

vmx_model_build_logs(run, client = vmx_client(), order = c("desc", "asc"))

Arguments

Value

A tibble containing all log lines in the requested order.

Description

Build-run report status (signed HTML report URL when ready)

Usage

vmx_model_build_report(run, client = vmx_client())

Arguments

Value

A list with status and, when ready, url.

Description

POST /model-build-runs/{run_id}/report queues HTML report generation.

Usage

vmx_model_build_report_create(
  run,
  subject_plot_mode = c("all", "none"),
  client = vmx_client()
)

Arguments

Value

A list with report status.

Description

Build-run results

Usage

vmx_model_build_results(run, client = vmx_client())

Arguments

Value

A list (fits summary, modeling population, PK structure selection).

Description

Run and fit reads request the API's marker-scopes representation, so runs that select a PD marker across all specimens are listed and readable, and PD selections report specimen_type or all_specimens.

Usage

vmx_model_build_runs(
  data_version = NULL,
  study = NULL,
  treatment = NULL,
  status = NULL,
  client = vmx_client()
)

Arguments

Value

A tibble containing all matching model-build runs.

Description

Build-run status

Usage

vmx_model_build_status(run, client = vmx_client())

Arguments

Value

A vmx_model_build_run.

Description

GET /model-catalog returns categories of models; this flattens them into one tibble with a category column.

Usage

vmx_model_catalog(data_version = NULL, client = vmx_client())

Arguments

Value

A tibble.

Description

Returns a vmx_model_data bundle with ⁠$subjects⁠, ⁠$pk⁠, ⁠$dosing⁠, ⁠$pd⁠, ⁠$labs⁠ and ⁠$covariates⁠ (each a tibble, or NULL when the DataVersion has no such prepared table), and ⁠$meta⁠ (units, time bases, the selected time_basis, PD-marker manifest, subject count) read from the DataVersion. Only domains flagged in the DV's table_availability are fetched, so absent optional tables don't 404. Every table is projected on the same time basis.

Usage

vmx_model_data(dv, time_basis = NULL, client = vmx_client())

Arguments

Value

A vmx_model_data object.

Description

Describe a model

Usage

vmx_model_describe(model_name, client = vmx_client())

Arguments

Value

A named list.

Description

Fetch one model fit's details

Usage

vmx_model_fit(id, client = vmx_client())

Arguments

Value

A vmx_model_fit (metadata / model / inference).

Description

Model-fit postprocessor status

Usage

vmx_model_fit_postprocessor_status(fit, client = vmx_client())

Arguments

Value

A list with postprocessor status.

Description

Requests the marker-scopes representation, so fits of a PD marker selected across all specimens are listed and each PD marker reports its specimen scope.

Usage

vmx_model_fits(
  run = NULL,
  data_version = NULL,
  model_type = NULL,
  marker_name = NULL,
  status = NULL,
  client = vmx_client(),
  treatment = NULL,
  study = NULL,
  source_pk_model_fit = NULL
)

Arguments

Value

A tibble containing all matching model fits.

Description

POST /modeling-options previews the modeling population and available covariates for one PK specimen scope, a set of PD marker selections and a time basis. Nothing is persisted.

Usage

vmx_modeling_options(
  data_version,
  time_basis,
  pk_specimen,
  pd_marker = NULL,
  covariate = NULL,
  client = vmx_client()
)

Arguments

Details

pk_specimen is required: one PK specimen type (for example "plasma") or vmx_all_specimens(). vmxr never fills it in, not even from the data version's recommendation; a call without it fails and lists the specimens available. Each PD marker is a vmx_pd_marker() naming one specimen or all specimens; any direction on it is ignored here, so the same selections can be passed on to vmx_model_build().

Value

A list (the selection preview).

Description

Creates the analysis (POST /nca-analyses) and, by default, blocks until it settles. time_basis is one of "observed", "nominal", "nominal_from_observed_dose", or "observed_from_predose_pk" (validated server-side against the DataVersion's available bases).

Usage

vmx_nca(
  data_version,
  time_basis,
  pk_specimen,
  idempotency_key = NULL,
  retried_from = NULL,
  wait = TRUE,
  ...,
  client = vmx_client()
)

Arguments

Details

pk_specimen is required: name one PK specimen (for example "plasma") or pass vmx_all_specimens() to pool every specimen. vmxr never chooses one for you, not even the data version's recommended specimen; a call without it fails and lists the specimens the data version offers. A named specimen needs the data version's per-specimen marker analysis. Read the accepted selection back with vmx_nca_marker_selection().

Value

A vmx_nca_analysis.

Description

List NCA analyses

Usage

vmx_nca_analyses(
  data_version = NULL,
  study = NULL,
  treatment = NULL,
  status = NULL,
  time_basis = NULL,
  client = vmx_client()
)

Arguments

Value

A tibble containing all matching analyses.

Description

Fetch one NCA analysis

Usage

vmx_nca_get(id, client = vmx_client())

Arguments

Value

A vmx_nca_analysis.

Description

Reads the NCA result and returns a tidy tibble of point estimates, handling both server shapes transparently:

Usage

vmx_nca_result(nca, client = vmx_client())

Arguments

Details

• API 0.3 returns a cursor-paged collection whose items[] are NcaResultItems — one per result interval of interest. Every page is followed to the end via next_cursor and the items are assembled without truncation. Each row is one subject within one interval, and the per-interval dimension is carried in the item_index, label, interval_start_hours, and interval_end_hours columns (the two ⁠*_hours⁠ columns are NA when the server did not resolve a time interval for that subject).

• API 0.2.2 returns a single flat object with top-level per-subject parallel arrays and one point_estimates map. Each row is one subject (there is no per-interval dimension, so the interval columns are absent).

In both shapes the point_estimates map (metric -> per-subject values, parallel to gen_subject_uuid) becomes one column per PK quantity, and not_estimable_reasons — the per-subject map of metric -> reason strings — is surfaced as the not_estimable_reasons list-column, so a null point estimate carries its reason. Quantity metadata (display names, units, explanations) is attached as the "quantities" attribute and the compact unit lookup as "units"; for 0.3 these are the de-duplicated union across items.

Value

A tibble.

Description

Assembles one event table from a DataVersion's prepared pk, dosing, subjects, covariates and (optionally) pd tables, in the layout nlmixr2 and rxode2 consume directly: one row per dose event and per observation, ordered by subject and time.

Usage

vmx_nlmixr_data(
  dv,
  analyte = NULL,
  time_basis = NULL,
  eligibility = c("after_qc", "before_qc", "all"),
  pd_markers = NULL,
  cmt = NULL,
  replicate = FALSE,
  client = vmx_client(),
  pk_specimen = NULL
)

Arguments

Details

Rows are ordered by subject and time; ties on TIME are ordered dose before observation, which is how rxode2 and nlmixr2 resolve them, so a pre-dose sample that shares its time with an IV bolus into the observation compartment must carry its own earlier time in the source data.

Observations that share a subject, endpoint and time are always emitted in a deterministic order — by the measurement uuid (gen_measurement_uuid, the row identity on the pk/pd tables) and then source-row order — whether or not the replicate index is requested. With replicate = TRUE the layout also carries a REPLICATE column: the 1-based rank of each observation within its ⁠(ID, DVID, TIME)⁠ group in that order, NA on dose rows and 1 for an observation with no same-time partner. The index is derived here in R and is deliberately not part of the DataVersion. rxode2 and nlmixr2 treat any extra column as a covariate, and REPLICATE is NA on dose rows, so drop it (dplyr::select(-REPLICATE)) before passing the table to a fit unless the model uses it explicitly.

Columns: ID (dense integer per subject with admitted rows), TIME (hours on the selected time basis), DV, AMT (mg), EVID (0 observation / 1 dose), MDV, CMT, RATE (mg/h, > 0 for infusions), II, ADDL, SS (always 0 — the DataVersion already expands doses per administration), CENS / LIMIT (censoring: CENS = 1 with DV = lloq and LIMIT = 0 for BLQ rows, CENS = -1 with DV = uloq and LIMIT = Inf for ALQ rows), DVID (only when PD markers are included: 1 = PK, 2.. = one per PD endpoint, and dose rows carry 1), subject_id, specimen_type (only when the DataVersion records specimens: each observation's own recorded specimen, also inside a pooled endpoint, and NA on dose rows), and one column per subject-level covariate.

A marker measured in more than one specimen (plasma and serum, say) is one marker with one input per specimen. The specimen is always the caller's choice: with more than one PK specimen, pk_specimen is required, and a PD marker with more than one specimen needs its specimens named in pd_markers. The DataVersion's recommended PK specimen is never used as an implicit choice. Each selection is admitted by the DataVersion's recorded eligibility verdict for that scope on the selected basis: the specimen's own verdict (by marker and specimen), or the All-specimens verdict when pooled. A scope verdict that is not on the DataVersion itself is read from its marker-analysis metadata. Pooling specimens that do not share one explicitly known unit is an error that names the units found. A DataVersion that predates specimens (no specimen_type column) is assembled exactly as before.

Only rows the DataVersion marks as eligible for modeling are kept; the after-QC flag is the contract's mandatory admission filter, and the result is analysis-ready only under that default. before_qc and all are review modes: they warn, mark the result analysis_ready = FALSE, and drop (with a count) any retained-ineligible row that cannot be encoded — a null time, a dose without canonical mass, a route outside the compartment map, an infusion without a canonical rate. Every dropped-row count is on the "vmx" attribute so nothing disappears silently.

Value

A tibble in NONMEM layout. The "vmx" attribute is a list with data_version_id, time_basis, units, analyte, pk_specimen (the accepted PK scope, list(specimen_type = ) or list(all_specimens = TRUE), or NULL when no specimen is recorded), eligibility, analysis_ready, cmt, endpoints (a tibble of dvid, name, cmt, unit, has_repeated_times — whether the endpoint carries any observation that shares its time — and, when the DataVersion records specimens, specimen_type (NA for a pooled endpoint) and all_specimens), dropped (named integer counts), and n_subjects. With replicate = TRUE the layout gains a REPLICATE column (see the argument).

See Also

vmx_model_data() for the underlying tidy tables.

Default route-to-compartment mapping for vmx_nlmixr_data()

Description

Extravascular routes (po, sc, im) dose into compartment 1 (depot); intravenous routes (iv-bolus, iv-infusion) dose into compartment 2 (central); PK observations are read from compartment 2. Override by passing a named integer vector with the same names to cmt.

Usage

vmx_nlmixr_default_cmt()

Description

PD observations table

Usage

vmx_pd(dv, time_basis = NULL, client = vmx_client())

Arguments

Value

A tibble.

Description

Builds one PD marker selection for vmx_modeling_options() and vmx_model_build(). Each selection names the marker's gen_uuid and either one specimen (for example "serum") or vmx_all_specimens(). Two specimens of the same marker may be selected as two separate entries; selecting one marker both in a named specimen and in all specimens is an error.

Usage

vmx_pd_marker(gen_uuid, specimen, direction = NULL)

Arguments

Value

A vmx_pd_marker object.

Examples

vmx_pd_marker("8ffb3792-8bc9-4d38-b123-fc3e2170f6ee", "serum", "decreasing")
vmx_pd_marker(
  "8ffb3792-8bc9-4d38-b123-fc3e2170f6ee", vmx_all_specimens(), "decreasing"
)

Description

PK observations table

Usage

vmx_pk(dv, time_basis = NULL, client = vmx_client())

Arguments

Value

A tibble.

Description

Answer prep questions and resume formatting

Usage

vmx_prep_answer(
  dataset,
  answers,
  client = vmx_client(),
  idempotency_key = NULL
)

Arguments

Description

Questions raised by prep (when awaiting input)

Usage

vmx_prep_questions(dataset, client = vmx_client())

Arguments

Value

A tibble of pending questions. Variable-shape values such as options, default, and data_preview are retained as list-columns.

Description

Calls GET /datasets/{ds_id}/prep-status.

Usage

vmx_prep_status(dataset, client = vmx_client())

Arguments

Value

A vmx_prep_status (status, and data_version_id once settled).

Description

Cancel a simulation job

Usage

vmx_sim_cancel(job, client = vmx_client())

Arguments

Value

A vmx_simulation_job.

Description

POST /model-fits/{mf_id}/existing-subject-simulation-jobs.

Usage

vmx_sim_existing_subject(
  fit,
  dosing_input,
  subjects,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

POST /model-fits/{mf_id}/existing-subject-simulation-jobs/from-text.

Usage

vmx_sim_existing_subject_from_text(
  fit,
  dosing_text,
  subjects,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

POST /model-fits/{mf_id}/hypothetical-subject-simulation-jobs.

Usage

vmx_sim_hypothetical_subject(
  fit,
  dosing_input,
  subjects,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

POST /model-fits/{mf_id}/hypothetical-subject-simulation-jobs/from-text.

Usage

vmx_sim_hypothetical_subject_from_text(
  fit,
  dosing_text,
  subjects,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

List simulation jobs for a model fit

Usage

vmx_sim_jobs(fit, client = vmx_client())

Arguments

Value

A tibble containing all simulation jobs for the fit.

Description

POST /model-fits/{mf_id}/population-simulation-jobs.

Usage

vmx_sim_population(
  fit,
  dosing_input,
  scenario_name,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

POST /model-fits/{mf_id}/population-simulation-jobs/from-text.

Usage

vmx_sim_population_from_text(
  fit,
  dosing_text,
  scenario_name,
  idempotency_key = NULL,
  retried_from = NULL,
  min_timepoints = NULL,
  wait = FALSE,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_simulation_job.

Description

GET /simulation-jobs/{id}/result. Returns the parsed result payload containing model-implied response trajectories with the server-provided point statistic and interval. The nested wire shape is retained verbatim.

Usage

vmx_sim_result(job, grouping_variable = NULL, client = vmx_client())

Arguments

Value

A list (the parsed result).

Description

Simulation job status

Usage

vmx_sim_status(job, client = vmx_client())

Arguments

Value

A vmx_simulation_job.

Description

List studies for a treatment

Usage

vmx_studies(
  treatment = NULL,
  status = NULL,
  client = vmx_client(),
  created_since = NULL
)

Arguments

Value

A tibble containing all matching studies.

Description

Fetch one study

Usage

vmx_study(id, client = vmx_client())

Arguments

Value

A vmx_study.

Description

Create a study

Usage

vmx_study_create(
  treatment,
  name,
  study_type = "clinical",
  phase = NULL,
  ...,
  client = vmx_client()
)

Arguments

Value

A vmx_study.

Description

Only the fields you pass are changed (the server applies exclude_unset).

Usage

vmx_study_update(id, ..., client = vmx_client())

Arguments

Value

A vmx_study.

Description

Subjects table (one row per subject)

Usage

vmx_subjects(dv, time_basis = NULL, client = vmx_client())

Arguments

Value

A tibble.

Description

Not yet implemented. The per-subject start[i]/end[i] index ranges and iObs observation index must be derived and verified against real data before shipping (this is the error-prone derivation the design flags); see the package NEWS. When implemented it should reuse the event assembly in vmx_nlmixr_data() (eligibility admission, basis selection, PK and PD specimen selection, dose and censoring encoding) and only add the ragged indexing on top.

Usage

vmx_torsten_data(dv, analyte = NULL, client = vmx_client())

Arguments

Value

Not yet implemented.

Description

Fetch one treatment

Usage

vmx_treatment(id, client = vmx_client())

Arguments

Value

A vmx_treatment.

Description

Create a treatment

Usage

vmx_treatment_create(
  name,
  indication = NULL,
  description = NULL,
  client = vmx_client()
)

Arguments

Value

A vmx_treatment.

Description

Only the fields you pass are changed (the server applies exclude_unset).

Usage

vmx_treatment_update(id, ..., client = vmx_client())

Arguments

Value

A vmx_treatment.

Description

List treatments

Usage

vmx_treatments(status = NULL, client = vmx_client())

Arguments

Value

A tibble containing all matching treatments.

Description

Streamed multipart upload. With wait = TRUE, blocks through the prep pipeline (see vmx_wait()).

Usage

vmx_upload(
  study,
  files,
  mode = c("initial", "incremental", "replacement"),
  treatment = NULL,
  config = NULL,
  wait = FALSE,
  client = vmx_client(),
  ...
)

Arguments

Value

A vmx_dataset (status "uploaded").

Description

POST /datasets/{ds_id}/ignore-upload. Soft-ignores one delivery.

Usage

vmx_upload_ignore(dataset, upload, client = vmx_client())

Arguments

Value

The updated vmx_prep_status.

Description

POST /datasets/{ds_id}/unignore-upload.

Usage

vmx_upload_unignore(dataset, upload, client = vmx_client())

Arguments

Value

The updated vmx_prep_status.

Description

An S3 generic dispatching on the handle type. Each method has a sensible default terminal state. Terminal-but-unsuccessful states raise a classed error so scripts fail loudly rather than hang.

Usage

vmx_wait(
  x,
  until = NULL,
  timeout = 900,
  interval = 5,
  progress = interactive(),
  client = vmx_client(),
  ...
)

Arguments

Value

The updated object, or a vmx_timeout_error.

Description

Calls GET /me.

Usage

vmx_whoami(client = vmx_client())

Arguments

Value

A vmx_me object: user_id, email, name, workspace_id, roles, and counts.