Import jsonlite::fromJSON()
#380
Comments
joshuaulrich
added a commit
that referenced
this issue
Apr 5, 2023
These are no longer needed since jsonlite is in Imports now. See #380.
joshuaulrich
added a commit
that referenced
this issue
Apr 5, 2023
The try() around jsonlite::fromJSON() is not necessary because the code is already inside a try(). This additional try() hides errors that may be thrown by the wrapped expression. This made it extremely hard to track down an error in a tseries example that called getSymbols.yahoo(). The call failed because tseries does not also Suggest jsonite, so jsonlite could not be found. See #380.
Closed
netbsd-srcmastr
pushed a commit
to NetBSD/pkgsrc
that referenced
this issue
Jun 12, 2023
### Changes in 0.4.22 (2023-04-05)
1. Move jsonlite from Suggests to Imports so it doesn't cause a problem
when a package that doesn't also Suggest jsonlite uses getSymbols().
Thanks to Kurt Hornik for the report and fix!
[#380](joshuaulrich/quantmod#380)
### Changes in 0.4.21 (2023-03-29)
1. Fix S3 method issues. R-devel (83995-ish) added a check for possible S3
method issues. Register methods it found that were not registered:
`str.replot()`, `seriesHi.timeSeries()`, and `seriesLo.timeSeries()`.
It was also confused by `range.bars()` and `unique.formula.names()`. Remove
`unique.formula.names()` because it wasn't exported or used internally.
Rename `range.bars()` to `rangeBars()`, which isn't exported.
Thanks to Kurt Hornik for the report!
[#375](joshuaulrich/quantmod#375)
1. Remove "^" prefix from `getSymbols()` return value. When the 'Symbols'
argument has a "^" prefix and `auto.assign = TRUE`:
* `getSymbols()` removes the "^" from the object it creates, but
* returns the 'Symbols' argument unchanged, and
* removes the "^" from the column names of the object it creates.
The example below will create an object named `IXIC` but the value of
`sym` will be "^IXIC".
sym <- getSymbols("^IXIC")
That means `x <- get(sym)` will not work because an object named `^IXIC`
doesn't exist.
[#371](joshuaulrich/quantmod#371)
1. Add 'from' and 'to' arguments to `getSymbols.FRED()`. Users expect to be
able to set the 'from' and 'to' arguments for FRED data like they can for
Yahoo data. Those values were ignored and the entire series was always
returned.
[#368](joshuaulrich/quantmod#368)
1. Change interval to 1d for `getDividends()` and `getSplits()`. The "3mo"
setting caused some dividends to be missing for companies that issued monthly
dividends. Note that the response to this request also includes all the OHLCV
data. But it's small (less than 1MB for 60+ years of daily data).
[#372](joshuaulrich/quantmod#372)
1. Handle errors in `getSplits()` and `getDividends()`. `getDividends()` didn't
handle cases where the download failed, or when dividends needed to be
split-adjusted but there were no splits. It also tried to set colnames
on the empty xts object that's returned when there are no dividends.
`getSplits()` had the same colnames issue. Check for no splits by testing
for `NULL` because that's more explicit. Thanks to Chris Cheung for the
report!
[#366](joshuaulrich/quantmod#366)
1. Export `HL()`, `is.HL()`, and `has.HL()` functions and add documentation.
These were added in 0.4.18 but not exported or included in the documentation.
1. Use Yahoo Finance v8 JSON endpoint and remove the v7 CSV endpoint. There
seems to be a rate limit for the number of tickers you can request via the CSV
endpoint. The [yfinance python library](https://github.com/ranaroussi/yfinance)
uses the JSON endpoint and doesn't seem to have rate limit issues.
[#360](joshuaulrich/quantmod#360)
[#362](joshuaulrich/quantmod#362)
[#364](joshuaulrich/quantmod#364)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Using
jsonlite::fromJSON()while jsonlite is only in Suggests causes a failure in the tseries package examples because tseries does not suggest jsonlite.Running
_R_CHECK_SUGGESTS_ONLY_=true R CMD check tseries --as-crancausesgetSymbols.yahoo()to fail.We need to import
jsonlite::fromJSON()to fix.Thanks to Kurt Hornik for the report and fix!
The text was updated successfully, but these errors were encountered: