bug fixes according to discussion

Author: RainerHeintzmann

docs/src/czt.md

@@ -1,5 +1,8 @@
 # CZTs
-Chirp Z Transformations:
+Chirp Z Transformations: Allows Fourier-transformation and at the same time zooming into the result,
+which is why it is also called the Zoomed-FFT algorithm.
+The algorithm is based on a publication [Rabiner, Schafer, Rader, The Cirp z-Transform Algorithm, IEEE Trans AU 17(1969) p. 86] and a matlab Version written by the group of H. Gross. It currently needs three FFTs to perform its work.
+As one of these FFTs only depends on the datasize and zoom parameters, it can be moved to a plan in future implementations.
 
 ```@docs
 FourierTools.czt

src/czt.jl

@@ -96,9 +96,11 @@ The tuple `scale` defines the zoom factors in the Fourier domain. Each has to be
 
 # Arguments:
 + `xin`: array to transform
-+ `scale`: a tuple of factors (one for each dimension) to zoom into during the czt. Note that a factor of nothing (or 1.0) needs to be provided, if a dimension is not transformed.
++ `scale`: a tuple of factors (one for each dimension) to zoom into during the czt. 
+   Note that a factor of nothing (or 1.0) needs to be provided, if a dimension is not transformed.
 + `dims`: a tuple of dimensions over which to apply the czt.
-+ `remove_wrap`: if true, the wrapped places will be set to zero. Note that the `pad_value` argument is only allowed for 1d czts to not cause confusion.
++ `remove_wrap`: if true, the wrapped places will be set to zero. 
+   Note that the `pad_value` argument is only allowed for czt_1d to not cause confusion.
 
 #Example:
 ```jdoctest
@@ -163,7 +165,8 @@ The tuple `scale` defines the zoom factors in the Fourier domain. Each has to be
 
 # Arguments:
 + `xin`: array to transform
-+ `scale`: a tuple of factors (one for each dimension) of the the inverse czt. Note that a factor of nothing (or 1.0) needs to be provided, if a dimension is not transformed.
++ `scale`: a tuple of factors (one for each dimension) of the the inverse czt. 
+   Note that a factor of nothing (or 1.0) needs to be provided, if a dimension is not transformed.
 + `dims`: a tuple of dimensions over which to apply the inverse czt.
 + `remove_wrap`: if true, the wrapped places will be set to zero. 
                  Note that the `pad_value` argument is only allowed for 1d czts to not cause confusion.

src/resampling.jl

@@ -138,7 +138,7 @@ By default the first pixel maintains its position. However, this leads to a shif
 `keep_singleton=true` will not upsample dimensions of size one.
 Note that upsample2 is based on Fourier-shifting and you may have to deal with wrap-around problems.
 
-```jdoctest
+```jldoctest
 julia> upsample2(collect(collect(1.0:9.0)'))
 2×18 Matrix{Float64}:
  1.0  0.24123  2.0  3.24123  3.0  2.93582  4.0  5.0  5.0  5.0  6.0  7.06418  7.0  6.75877  8.0  9.75877  9.0  5.0
@@ -190,7 +190,7 @@ per entry along this dimension number.
 + `new_size`: size of the result array. If not provided the same as the input size will be used.
 
 # Examples
-```jdoctest
+```jldoctest
 julia> using TestImages, NDTools, View5D
 
 julia> a = Float32.(testimage("resolution"));
@@ -297,7 +297,7 @@ using `resample_czt()` with separable quadratic zooms.
 
 See also: `resample_czt()`
 # Examples
-```jdoctest
+```jldoctest
 julia> using TestImages, NDTools, View5D
 
 julia> a = Float32.(testimage("resolution"))
@@ -344,7 +344,7 @@ the result has usually less artefacts, but the positions may be more less conven
 
 See also: `resample`, `resample_czt`
 # Examples
-```jdoctest
+```jldoctest
 julia> using FourierTools, TestImages, NDTools, View5D, IndexFunArrays
 
 julia> a = Float32.(testimage("resolution"));