exploratory-io / exploratory_func Goto Github PK

Issue Description

Label needs to be updated from Current Year to This Year, Current Month to This Month, Current Week to This Week.

Issue Description

If new data to predict doesn't have the response column used to build xgboost_binary, this error appears

model frame and formula mismatch in model.matrix()

Steps to Reproduce

Provide detail steps to reproduce the issue, including which data you use

Expected Behavior

What should happen

Actual Behavior

What really happened

Error Message

Provide error message or screenshot if available

Other Comments

I want to run the do_cosine_sim.kv() function for comparing two corpuses, but when loading the exploratory package I run in to an error that R can't find the function.

> library(exploratory)
Package attached: exploratory v0.3.16 (same as the most recent version available through GitHub).

If you run into errors, please try restarting R.

> do_cosine_sim.kv()
Error in do_cosine_sim.kv() : could not find function "do_cosine_sim.kv"

All help much appreciated.

Issue Description

devtools::check() fails because of insufficient dependency information

I see this error

R CMD check results
1 error  | 0 warnings | 0 notes
checking package dependencies ... ERROR
Package suggested but not available for checking: ‘gtrendsR’

Namespace dependencies not required:
  ‘data.table’ ‘dtw’ ‘foreach’ ‘ggplot2’ ‘iterators’ ‘scales’ ‘zoo’

See section ‘The DESCRIPTION file’ in the ‘Writing R Extensions’
manual.
Warning message:
do_market_impact_.Rd is missing name/title. Skipping 

install from github also fails because of this

Steps to Reproduce

Provide detail steps to reproduce the issue, including which data you use

Expected Behavior

What should happen

Actual Behavior

What really happened

Error Message

Provide error message or screenshot if available

Other Comments

countrycode function should be faster to use it in colopreth map of Exploratory desktop

Now, we use https://github.com/vincentarelbundock/countrycode directly.

You can try

• stringr functions instead of default functions like gsub
• Rcpp or RcppParallel
  • http://adv-r.had.co.nz/Rcpp.html
  • https://rcppcore.github.io/RcppParallel/

Todo:

• Create examples for benchmark
• Check benchmark of current countrycode function
• Implement faster function
• Compare the benchmark

It should return result immediately, not model

I am trying to reproduce the example from the manual

> res <- data.frame("text" = c("this is what it is", "which is better")) %>%
+   do_tokenize(text) %>%
+   do_tfidf(document_id, token)

which is expected to result in:

However, I obtain

Another strange result is the following:

> data.frame("text" = c("good it was", "is nice she", "good is she")) %>%
+   do_tokenize(text) %>%
+   do_tfidf(document_id,token)

where I would expect to find identical values for "it" and "was"...

Issue Description

If you run do_anomaly_detection to data with non-UTC timezone, it fails to find anomaly data.

Steps to Reproduce

Provide detail steps to reproduce the issue, including which data you use

Expected Behavior

What should happen

Actual Behavior

What really happened

Error Message

Provide error message or screenshot if available

Other Comments

Issue Description

Currently, scatter plots for coefficients of Linear regression etc. are sorted based on coefficient values of the first facet.
With that default behavior, we could provide options to...

• Sort within each facet based on coefficient values
• No sort. All facets are with alphabetical order of predictor names.

do_cosine_sim.kv doesn’t work under [email protected]. Running the function with a dataset that should work fine (a bit tricky to share, but can if need be—just creating the issue to note this is a problem) generates the following:

Error: `unnest_()` is deprecated as of tidyr 1.0.0.
Please use `unnest()` instead. 

Issue Description

Like this screenshot, reference Line with constant value is not displayed when X-axis has only 1 value on scatter plot.

Issue Description

tidyverse/readr@cc16a45

to align with the change, need to use Inf as default value for n_max argument.

Trying to install exploratory package in R 3.6

ERROR: dependency 'anonymizer' is not available for package 'exploratory'
* removing 'C:/Users/proctors/Documents/R/win-library/3.6/exploratory'
Error: Failed to install 'exploratory' from GitHub:

Seems like it was removed?
https://cran.r-project.org/web/packages/anonymizer/index.html

Issue Description

Currently, list_extract supports position for extract target
And there is a requirement for support boolean logic like position >= 4 to extract a value.