Fix error related to filter(.testdata, spc > 100)

[GegznaV]

Code that worked previously now results in an error:

filter(.testdata, spc > 100)

May be related to the recent changes (up to 1-2 years) in Tidyverse.

@eoduniyi, please take care of this.

Related:

• Fix some R-CMD-check errors #50

[eoduniyi]

Code that worked previously now results in an error:

filter(.testdata, spc > 100)

May be related to the recent changes (up to 1-2 years) in Tidyverse.

@eoduniyi, please take care of this.

Related:

• Fix some R-CMD-check errors #50

According to dplyr 1.0.8, filter() only allows matrices with a single column.

[eoduniyi]

@GegznaV or @cbeleites could you please give some guidance on the default behavior of "filtering" the spectral matrix?:

Before the dplyr 1.0.8 update, filter(.testdata, spc > 100) would produce a hyperSpec object that would only keep the spectra whose intensity values were all > 100.

Alternatively, users could specify to keep the spectra if any of the spectrum's intensity were > 100

However, it's not clear to me where all_wl() or any_wl() is being applied in filter.hyperSpec()'s definition.

[eoduniyi]

@GegznaV or @cbeleites could you please give some guidance on the default behavior of "filtering" the spectral matrix?:

Before the dplyr 1.0.8 update, filter(.testdata, spc > 100) would produce a hyperSpec object that would only keep the spectra whose intensity values were all > 100.

Alternatively, users could specify to keep the spectra if any of the spectrum's intensity were > 100

However, it's not clear to me where all_wl() or any_wl() is being applied in filter.hyperSpec()'s definition.

🤦🏿‍♂️ filter(.testdata, any_wl(spc > 100)) and filter(.testdata, all_wl(spc > 100)) work as expected.

[GegznaV]

In the Tidyverse, filter() keeps/removes rows based on provided conditions. The length of the logical vector that is created based on conditions should be equal to nrow(.data). So filter(.data, spc > 100) seems ambiguous to me, as the intensity at some wavelengths might fulfill the condition and at some others might not for the same row. So I'd also expect the failure when a condition such as spc > 100 is used.