new-in32.txt 9.9 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230
  1. This file is part of GRG 3.2 Copyright (C) 1997 Vadim V. Zhytnikov
  2. This note outlines the main new features of GRG 3.2 compared to
  3. the previous version GRG 3.1. Detailed description of GRG 3.2 can
  4. be found in the manual.
  5. 1. Dimensionality and Signature
  6. GRG 3.2 removes the most important restriction of GRG 3.1 -
  7. now GRG 3.2 works in any dimension greater or equal 2.
  8. The only signatures GRG 3.1 works with are (-1,1,1,1) or
  9. (1,-1,-1,-1). In GRG 3.2 you can specify any signature of the
  10. form diag(+1,-1,...).
  11. 2. Metric and Frame
  12. In GRG 3.1 the space-time metric can be specified only in the form
  13. of the tetrad and the tetrad metric must be quasi-orthogonal (here
  14. "quasi" means any metric having constant coefficients: null,
  15. semi-null, diagonal Lorentzian etc). In GRG 3.2 you can use
  16. arbitrary metric and frame (in GRG 3.2 we use the name "frame"
  17. instead of "tetrad" since the latter is essentially 4-dimensional).
  18. In general the space-time line-element in GRG 3.1 and 3.2 is
  19. 2 i j
  20. d s = G T T
  21. ij
  22. Here G.i.j (M in GRG 3.1) is the Metric and T'i is frame (former
  23. Tetrad). In GRG 3.1 the G (M) has predefined value equals to Null
  24. Metric and can be changed only by assigning a new value to M or
  25. loading the file "lorentz.loa" or similar. Even more, the coefficients
  26. of M in GRG 3.1 must be constant ("tetrad" formalism). In GRG 3.2 both
  27. G.i.j and T'i are initially indefinite. You can assign any value
  28. to these quantities and G is no not necessary constant now. But if
  29. no value is given to G or T then GRG 3.2 automatically assumes
  30. the default values to these quantities
  31. i i
  32. T = d x
  33. G = diag(i,j)
  34. ij
  35. where diag(i,j) is the matrix having only diagonal nonzero
  36. elements +1 or -1 according to current signature. For example,
  37. if dimensionality 3 with the signature diag(+1,-1,+1) and no
  38. value is specified to T and G than GRG 3.2 automatically assumes
  39. 0 1 2
  40. G = diag(+1,-1,+1), T = d x0, T = d x1, T = d x2,
  41. ij
  42. which gives finally
  43. 2 2 2 2
  44. d s = d x0 - d x1 + d x2
  45. where x0, x1, x2 are the coordinates. Thus, if you want
  46. to work in "tetrad" formalism - give the value to T and
  47. leave G to be default constant diagonal metric. Otherwise,
  48. if you want to use usual coordinate formalism - assign the
  49. value to G only and the default holonomic value will be
  50. automatically assigned to frame T.
  51. 3. Nonmetricity
  52. GRG 3.2 works with arbitrary affine connection having both the
  53. torsion and nonmetricity. The corresponding switches TORSION and
  54. NONMETR determine the connection type.
  55. 4. Better Representation of Built-in and User-Defined objects
  56. GRG 3.2 knows various symmetries with respect to index permutation:
  57. symmetry, antisymmetry, cyclic symmetry and Hermitian symmetry.
  58. All these symmetries can be applied to both single indices and the
  59. groups of indices. The groups of symmetric indices can be nested
  60. (like the symmetries of the Riemann curvature tensor).
  61. In GRG 3.2 the built-in and user-defined objects can have indices
  62. of the following types: holonomic (coordinate) indices (new),
  63. frame indices, undotted and dotted spinorial indices, and
  64. enumerating indices. All indices can be upper and lower (upper
  65. spinorial indices are new).
  66. GRG 3.2 understands tensor densities and pseudo-tensors.
  67. The "New Object" declaration is redesigned to account for the
  68. aforementioned changes.
  69. 5. Output Modes
  70. The new and quite unique feature of GRG 32 is that it can export
  71. the results of computations into other programs. Now GRG 3.2
  72. supports output modes for all major computer algebra systems
  73. Maple, Mathematica, Macsyma, REDUCE and the document preparation
  74. system LaTeX. Thus, for example, you can write result of
  75. computation into a file in the Maple input format and later use
  76. Maple to work with the data. LaTeX output mode allows one to
  77. insert the results of computation into a document.
  78. 6. GRG 3.2 is compatible with the REDUCE graphic-shells such as XR
  79. (under UNIX) or PSLLW (under MS Windows). In graphic mode GRG 3.2
  80. provides niece output with greek characters, integral signs etc.
  81. 7. The quite restrictive feature of GRG 3.1 is that it permits to use
  82. only some fixed set of the REDUCE built-in functions (such as SIN,
  83. COS, LOG etc) and switches. GRG 3.2 is not so restrictive.
  84. You can use any REDUCE switch, function (operator in the REDUCE
  85. terminology) or user-defined procedure. For example:
  86. (a) You can load the package specfn and use in GRG all special
  87. function defined in this package.
  88. (b) You can write your own procedure in the REDUCE language
  89. and apply it in GRG.
  90. (c) You can use built-in REDUCE operators such as INT, LIMIT,
  91. SUM, PROD, DEN, NUM and any other.
  92. Notice also GRG 3.2 allows one to apply REDUCE procedures and
  93. functions not only to algebraic expressions but to vectors
  94. and exterior forms in accordance to the distributive law.
  95. 8. Many GRG built-in objects and formulas (ways of calculation) have
  96. limited scope. Some are valid in a particular dimension only,
  97. some, like spinors, require standard null frame etc. GRG 3.2
  98. always checks the applicability conditions for any built-in
  99. object and formula.
  100. 9. The mechanism for computing covariant differentials and Lie
  101. derivatives in GRG 3.1 is rather clumsy. In GRG 3.2 it is
  102. completely replaced by another simpler and more natural method:
  103. one can use covariant exterior differentials, covariant
  104. derivatives and Lie derivatives directly in expressions.
  105. 10. In GRG 3.2 one can access the left- and right-hand side of
  106. equations using LHS(...) and RHS(...). The equations are
  107. properly transformed under the frame (tetrad) rotation and
  108. the coordinate transformations.
  109. 11. In GRG 3.2 expressions are case sensitive but commands and names
  110. of objects are not. Thus, the variable alpha is different from
  111. ALPHA but command Find is the same as find, FIND etc.
  112. 12. Usually irreducible spinors are labelled in GRG by so called
  113. summed spinorial index. This method provides the most efficient
  114. way to store irreducible spinor components but it is inconvenient
  115. when the spinor is used in expressions. Now any irreducible
  116. spinor can be labeled by both single summed index of rank N or by
  117. a group of N single spinorial indices. Analogously the frame
  118. index (for null frame) can be also represented as a pair of
  119. single spinorial indices.
  120. 13. GRG 3.2 provides simple interface to the REDUCE algebraic
  121. equation solver. The solutions can be used in the "Let" and "SUB"
  122. substitutions.
  123. 14. GRG 3.2 includes algebraic classification schemes for
  124. the following spinors: Weyl spinor W_ABCD, Traceless
  125. Ricci spinor C_AB_CD~, Electromagnetic stress spinor F_AB,
  126. Vector V_A_B~ (in spinorial representation). The work
  127. of all classification algorithms is traced. The command
  128. Petrov Type; is replaced by the command Classify <object>;.
  129. 15. If some built-in object has several ways of calculation GRG 3.2
  130. provides better method for choosing the particular way. One can
  131. indicate the way both by its name and by specifying any object
  132. which is present in the right-hand side of the corresponding
  133. formula.
  134. 16. GRG 3.2 provide interface to the REDUCE package dfpart.red
  135. written by H. Melenk. This very useful package introduces the
  136. notion of the partial derivative of a function with respect to
  137. its n-th argument and performs the chain differentiation.
  138. 17. Configuring GRG
  139. GRG 3.2 has some configuration facilities. The configuration
  140. file allows one to define:
  141. (a) the default dimensionality and signature,
  142. (b) the default position of switches,
  143. (c) the packages which must be preloaded,
  144. (d) the synonymy for the commands.
  145. Actually there are two configuration files. The first grgcfg.sl
  146. defines the "global" GRG configuration at the moment of the
  147. compilation. You can edit this file before compiling GRG and the
  148. corresponding settings will be active whenever GRG is started.
  149. The second configuration file grg.cfg is optional. You can keep
  150. it in your working directory to override the "global" settings.
  151. Another configuration tool is the environment variable "grg".
  152. This environment variable should contain the name of some
  153. directory (so called GRG System Directory). This directory can
  154. serve as the depository for the files which are oftenly used.
  155. So it is not necessary to have their copy in every working
  156. directory.
  157. 18. The GRG commands are terminated now only by the symbol ;
  158. (not by ; and ?). The end-of-file symbol for GRG batch
  159. files has changed from ! to $.
  160. 19. The GRG 3.1 command Help <object>; is replaced by the
  161. command Show <object>;. The commands Mode; and Signature;
  162. are replaced by the command Status;.
  163. 20. Built in Objects and Ways of calculation
  164. The list of built-in objects and formulas is significantly
  165. changed. Some objects and ways of calculation are renamed,
  166. some abolished but many other are added. Now GRG 3.2 has
  167. built-in object and formulas for:
  168. - Connection, torsion, nonmetricity (new).
  169. - Curvature.
  170. - Irreducible decomposition of the curvature, torsion, and
  171. nonmetricity in any dimension (new).
  172. - Einstein equations.
  173. - Scalar field with minimal and non-minimal interaction.
  174. - Electromagnetic field.
  175. - Yang-Mills field.
  176. - Dirac spinor field.
  177. - Geodesic equation (new).
  178. - Optical scalars (new).
  179. - Kinematics for time-like congruences (new).
  180. - Ideal fluid and spin fluid (new).
  181. - Newman-Penrose formalism (new).
  182. - Gravitational equations for the theory with arbitrary
  183. gravitational Lagrangian in Riemann and Riemann-Cartan
  184. spaces.
  185. ----------------------------------------------------------------------