Differences

This shows you the differences between two versions of the page.

Link to this comparison view

documentation:ref:generatetensor [2015/11/21 12:33]
 documentation:ref:generatetensor [2015/11/21 12:33] (current)
Line 1: Line 1:
 +====== GenerateTensor ======
 +-----
 +
 +====Description====
 +  * ''​GenerateTensor(indices,​ samples)''​ generates tensor of the most general form from a given samples. ​
 +
 +  * The [[SimpleIndices|indices]] of tensor produced by ''​GenerateTensor(indices,​ samples)''​ are ''​indices''​.
 +
 +  * The [[documentation:​guide:​symmetries_of_tensors|symmetries]] of tensor produced by ''​GenerateTensor(indices,​ samples)''​ will be equal to symmetries of specified [[simpleindices]] ''​indices''​ .
 +
 +  * ''​GenerateTensor(indices,​ samples, options)''​ allows to pass additional [[#​Options|options]]. ​
 +
 +====Examples====
 + The most general tensor with 3 indices that can be assembled from metric tensor $g_{mn}$ and vector $k_m$ is
 +$$
 +c_1 k_a k_b k_c + c_2 g_{ac} k_b + c_3 g_{ab} k_c + c_4 g_{bc} k_a
 +$$
 +With Redberry one can do the following
 +
 +<sxh groovy; gutter: false>
 +def t = GenerateTensor('​_abc'​.si,​ ['​g_mn',​ '​k_a'​])
 +println t
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > C[0]*k_{a}*k_{b}*k_{c}+C[1]*g_{ac}*k_{b}+C[2]*g_{ab}*k_{c}+C[4]*g_{bc}*k_{a}
 +</​sxh>​
 +
 +----
 +
 +Generate tensor with 4 indices with particular symmetries:
 +<sxh groovy; gutter: false>
 +def indices = '​_{abcd}'​.si //parse SimpleIndices
 +indices.symmetries.add(-[[0,​ 2, 1, 3]].p) //add particular symmetries
 +t = GenerateTensor(indices,​ ['​g_ab',​ '​k_a'​])
 +println Collect['​C[x]'​.t] >> t
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > C[0]*(g_{ab}*k_{c}*k_{d}-g_{cd}*k_{a}*k_{b})
 +         ​+C[1]*(-g_{bc}*k_{a}*k_{d}-g_{ad}*k_{b}*k_{c}
 +                +g_{bd}*k_{a}*k_{c}+g_{ac}*k_{b}*k_{d})
 +         ​+C[2]*(g_{ac}*g_{bd}-g_{ad}*g_{bc})
 +</​sxh>​
 +
 +----
 +Generate fully antisymmetric tensor with 5 indices from samples ''​t_mn''​ and ''​f_abc'':​
 +<sxh groovy; gutter: true>
 +def indices = '​_abcde'​.si
 +indices.symmetries.setAntiSymmetric()
 +def r = Collect['​C[x]'​] >> GenerateTensor(indices,​ ['​t_mn',​ '​f_abe'​].t)
 +println r
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > C[0]*(-t_{be}*f_{dca}+t_{da}*f_{ecb}-t_{ba}*f_{ecd}+t_{bc}*f_{ade}
 +          +t_{eb}*f_{dca}-t_{ce}*f_{abd}-t_{ea}*f_{cbd}-t_{da}*f_{ceb}
 +          +t_{ca}*f_{ebd}+t_{ba}*f_{edc}+t_{ed}*f_{acb}-t_{ca}*f_{dbe}
 +          -t_{ba}*f_{dec}-t_{be}*f_{adc}+t_{eb}*f_{adc}+t_{de}*f_{bca}
 +          -t_{ec}*f_{adb}+t_{dc}*f_{eba}-t_{ab}*f_{ced}-t_{ad}*f_{bec}
 +          -t_{da}*f_{ebc}+t_{ea}*f_{cdb}+t_{be}*f_{dac}-t_{dc}*f_{abe}
 +          -t_{eb}*f_{dac}+t_{ae}*f_{dcb}-t_{bd}*f_{cea}-t_{ae}*f_{dbc}
 +          -t_{ea}*f_{bdc}+t_{dc}*f_{bae}-t_{db}*f_{cae}-t_{dc}*f_{bea}
 +         ​+t_{cb}*f_{eda}+t_{ab}*f_{cde}-t_{ad}*f_{ecb}-t_{ac}*f_{deb}
 +         ​+t_{da}*f_{cbe}-t_{ce}*f_{bda}+t_{ec}*f_{abd}+t_{ac}*f_{bed}
 +         ​+t_{ad}*f_{ceb}-t_{ab}*f_{dce}+t_{ac}*f_{edb}-t_{dc}*f_{eab}
 +         ​-t_{da}*f_{bce}+t_{bd}*f_{aec}-t_{de}*f_{acb}+t_{ab}*f_{ecd}
 +          +t_{ce}*f_{dba}+t_{ae}*f_{cbd}-t_{bd}*f_{ace}-t_{ab}*f_{edc}
 +          -t_{cb}*f_{ead}-t_{cb}*f_{dea}+t_{ad}*f_{ebc}-t_{be}*f_{cad}
 +          +t_{dc}*f_{aeb}-t_{bc}*f_{eda}-t_{bd}*f_{eac}+t_{ab}*f_{dec}
 +          +t_{cb}*f_{dae}+t_{ce}*f_{bad}+t_{eb}*f_{cad}+t_{ea}*f_{bcd}
 +          +t_{ca}*f_{deb}+t_{be}*f_{acd}-t_{ca}*f_{bed}-t_{ae}*f_{cdb}
 +          -t_{eb}*f_{acd}-t_{ac}*f_{bde}+t_{cb}*f_{aed}-t_{ca}*f_{edb}
 +          -t_{ad}*f_{cbe}+t_{ed}*f_{cba}+t_{bd}*f_{eca}+t_{ec}*f_{bda}
 +          -t_{cd}*f_{eba}+t_{ae}*f_{bdc}+t_{db}*f_{cea}-t_{ed}*f_{abc}
 +          +t_{bc}*f_{ead}+t_{bc}*f_{dea}-t_{ed}*f_{cab}+t_{ad}*f_{bce}
 +          -t_{ce}*f_{dab}+t_{cd}*f_{abe}+t_{ed}*f_{bac}-t_{bc}*f_{dae}
 +          -t_{ec}*f_{dba}-t_{cd}*f_{bae}+t_{cd}*f_{bea}-t_{db}*f_{aec}
 +          +t_{ca}*f_{bde}-t_{ec}*f_{bad}-t_{bc}*f_{aed}+t_{db}*f_{ace}
 +          +t_{ba}*f_{ced}+t_{cd}*f_{eab}+t_{be}*f_{cda}-t_{cb}*f_{ade}
 +          -t_{ae}*f_{bcd}+t_{db}*f_{eac}-t_{eb}*f_{cda}-t_{ed}*f_{bca}
 +          +t_{ce}*f_{adb}-t_{ac}*f_{ebd}-t_{de}*f_{cba}-t_{ea}*f_{dcb}
 +          +t_{da}*f_{bec}+t_{ea}*f_{dbc}+t_{bd}*f_{cae}+t_{ac}*f_{dbe}
 +          +t_{de}*f_{abc}-t_{cd}*f_{aeb}-t_{ba}*f_{cde}+t_{ec}*f_{dab}
 +          +t_{de}*f_{cab}-t_{db}*f_{eca}-t_{de}*f_{bac}+t_{ba}*f_{dce})
 +</​sxh>​
 +Check its antisymmetry property:
 +<sxh groovy; gutter: true; first-line: 6>
 +def expr = "​F_abcde = $r".t
 +println expr >> '​F_abcde + F_abdce'​.t
 +</​sxh>​
 +<sxh plain; gutter: false>
 +    > 0
 +</​sxh>​
 +<sxh groovy; gutter: true; first-line: 8>
 +println expr >> '​F_abcde + F_decba'​.t
 +</​sxh>​
 +<sxh plain; gutter: false>
 +    > 0
 +</​sxh>​
 +====Options====
 +  * ''​SymmetricForm'':​\\ produces completely symmetric tensor:<​sxh groovy; gutter: false>
 +println GenerateTensor('​_{abc}'​.si,​ ['​g_mn',​ '​k_m'​],​ [SymmetricForm:​ true])
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > (1/​3)*C[0]*(g_{bc}*k_{a}+g_{ac}*k_{b}+g_{ab}*k_{c})+C[1]*k_{a}*k_{b}*k_{c}
 +</​sxh>​
 + 
 +  * ''​GeneratedParameters'':​ \\ Allows to control how free parameters are generated:<​sxh groovy; gutter: false>
 +println GenerateTensor('​_{ab}'​.si,​ ['​g_mn',​ '​k_m'​])
 +</​sxh><​sxh plain; gutter: false>
 +   > C[0]*g_ab+C[1]*k_{a}*k_{b}
 +</​sxh>​Control generated parameters: <sxh groovy; gutter: false>
 +println GenerateTensor('​_{ab}'​.si,​['​g_mn',​ '​k_m'​], ​
 +    [GeneratedParameters:​ {i -> "​K$i"​.t}])
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > K0*g_ab+K1*k_{a}*k_{b}
 +</​sxh>​
 +
 +  * ''​GenerateParameters'':​ \\ if false then no parameters will be generated:<​sxh groovy; gutter: false>
 +println GenerateTensor('​_{ab}'​.si,​ ['​g_mn',​ '​k_m'​],​ [GenerateParameters:​ false])
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > g_ab+k_{a}*k_{b}
 +</​sxh>​
 +
 +
 +  * ''​RaiseLower'':​ \\ By default ''​GenerateTensor''​ tries to upper/lower indices of the provided samples to generate tensor of the most general form. If ''​RaiseLower''​ set to false, then it will use samples as is: <sxh groovy; gutter: false>
 +println GenerateTensor('​_{ab}^{cd}'​.si, ​ ['​g_mn',​ '​k_m',​ '​k^m'​],​ [RaiseLower:​ false])
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > C[0]*g_{ab}*k^{c}*k^{d}+C[1]*k_{a}*k_{b}*k^{c}*k^{d}
 +</​sxh> ​
 +<sxh groovy; gutter: false>
 +println GenerateTensor('​_{ab}^{cd}'​.si,​ ['​g_mn',​ '​k_m'​]).size()
 +</​sxh>​
 +<sxh plain; gutter: false>
 +   > 10
 +</​sxh>​
 +
 +====See also====
 +  * Related guides: [[documentation:​guide:​tensors_and_indices]],​ [[documentation:​guide:​symmetries_of_tensors]]
 +  * Related reference material: [[documentation:​ref:​reduce]],​ [[documentation:​ref:​simpleindices]] ​
 +  * JavaDocs: [[http://​api.redberry.cc/​redberry/​1.1.9/​java-api/​cc/​redberry/​core/​tensorgenerator/​TensorGenerator.html| TensorGenerator]]
 +  * Source code: [[https://​bitbucket.org/​redberry/​redberry/​src/​tip/​core/​src/​main/​java/​cc/​redberry/​core/​tensorgenerator/​TensorGenerator.java|TensorGenerator.java]]