NEWS.md

In robertzk/cachemeifyoucan: Cache Me If You Can

Version 0.2.4.6

• If the blacklist contains NA, only prevent caching rows that have all missing values.

Version 0.2.4.5

• Fixed a few places where we should have used drop = FALSE.

Version 0.2.4.4

• Fixed a bug when running a 1x1 dataframe through a blacklist.

Version 0.2.4.3

• Fix a typo that would result in a crash if dplyr::bind_rows didn't work (instead of handing over to rbind.fill as intended).

Version 0.2.4.2

• Add Remotes to DESCRIPTION.

Version 0.2.4.1

• Re-replaces dplyr::bind_rows with plyr::rbind.fill because plyr::rbind.fill allows the combination of dataframes with alike content but different classes (e.g., "FALSE" and FALSE), which happens due to type coercion the cached data.

Version 0.2.4

• Implements a blacklist of values that are to be returned but not cached.
• Clean up how safe_columns is passed along in the function environment.

Version 0.2.3.6

• Fix last_cached_at bug for existing shards.

Version 0.2.3.5

• Fix safe_column logic to do what it actually intended.

Version 0.2.3.4

• Add last_cached_at column in each shard.

Version 0.2.3.3

• Update the update_cache_salt utility to be more proof.

Version 0.2.3.2

• Add a utility to update cache salts.

Version 0.2.3.1

• Switched to using dbtest::db_test_that from the latest version of dbtest.

Version 0.2.3

• Adds safe_columns, which, when TRUE will crash when (a) column tables already exist for a given cache and (b) new columns are found and added.
• Restructures the codebase to allow this hard crash to not be discarded by the myriad of try blocks.

Version 0.2.2

• Moved the db testing infrastructure to another package, syberia/dbtest.

Version 0.2.1.4-6

• Use batchman parallelism and add dry. parameter to cache.

Version 0.2.1.3

• Using double precision types during appends, as well as correctly setting database types upon table creation due to RPostgres bug.

Version 0.2.1.2

• Numeric values were being recorded using the real data type, which only supports 6 digits of precision. This was fixed by switching these column types to the numeric data type.

Version 0.2.1.1

• Restore column names from existing cache db when using force..

Version 0.2.1

• Debug mode, triggered by cachemeifyoucan.debug option. Set it to true to see the gory metadata internals of underlying postgres tables.

Version 0.2.0.10

• Ensure uniqueness of the response of the caching layer.

Version 0.2.0.9

• On second thought, don't store the shards in legacy cache warnings in globalenv for debugging, since they could be huge.

Version 0.2.0.8

• Introduce some changes to the merge2 helper that will ensure legacy caches that have discrepancy between shards will not truncate their output.

Version 0.2.0.7

• Added cache migrations and slightly refactored how new shards are created.

Version 0.2.0.6

• Important bugfix for sharding.

Version 0.2.0.5

• Replace RPostgreSQL with rstats-db/RPostgres.

Version 0.2.0.4

• Fix an issue with invalid index names.

Version 0.2.0.3

• Fixed some bugs and added new ones.

Version 0.2.0.2

• Re-define uncached as simply stripping the caching layer, after some confusion as to what the correct definition should be.

Version 0.2.0.1

• Fixed a bug in uncached.

Version 0.2.0

• Added column sharding, which allows to store data frames with large number of columns.
• Rocco documentation.
• Automatically create indexes on shards, which greatly improves read times.

Version 0.1.7

• Integrated batchman v1.0.0.9000, which fixes robust functionality.

Version 0.1.6.2

• If two users are populating the caching layer at the same time, the uncached function will not be run twice by both functions. Instead, each attempt to run the uncached function will query the database for whether those records have been cached. This leads to a speed-up during parallel cache population.

Version 0.1.6.1

• Added cachemeifyoucan.verbose global option for whether to display batch caching progress.

Version 0.1.6

• Batchman integration now uses robust_batch.

Version 0.1.5

• Simple generic bug fix.

Version 0.1.4

• Critical fix for usage of the force. parameter.

Version 0.1.3

• Integration with batchman. If a call to a cached function procures uncached data of row count in excess that of the parameter batch_size to the cachemeifyoucan::cache call, caching progress will automatically happen in batches of batch_size (by default 100). In other words, data will be cached on a batch_size record by record basis, instead of all at once, which is very susceptible to failures.

  Note this feature will only trigger if the batchman package is installed.

Version 0.1.2

• Bug fix release.

Version 0.1.1

• Better support and fixes for MonetDB adapter.

Version 0.1.0

• Initial creation of the package.

robertzk/cachemeifyoucan documentation built on May 27, 2019, 10:34 a.m.