{
  "id": "null-space",
  "title": "Null Space",
  "lang": "en",
  "source_lang": "en",
  "summary": "The null space of a linear map is the vector subspace of inputs mapped to zero, describing homogeneous solutions and nonuniqueness.",
  "category": "mathematics",
  "subcategory": "algebra",
  "type": "concept",
  "aliases": [
    "Nullspace",
    "Kernel",
    "Null space of a matrix"
  ],
  "generated_at": "2026-10-04",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/null-space/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/null-space/",
      "json": "https://aiwiki.page/zh/null-space.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "In linear algebra, the null space of a linear map is the set of vectors that it maps to the zero vector. Also called its kernel, it describes the inputs that the map annihilates. For a matrix \\(A\\), the null space consists of all solutions of the homogeneous equation \\(Ax=0\\). It is a linear subspace of the input space, not generally of the output space. (maths.tcd.ie)"
    },
    {
      "heading": "Definition and basic properties",
      "text": "Let \\(T:V\\to W\\) be a linear map between vector spaces over a field \\(\\mathbb F\\). Its null space is\n\n\\[\\ker T=\\{v\\in V:T(v)=0_W\\}.\\]\n\nFor an \\(m\\times n\\) matrix over \\(\\mathbb F\\), common notations include\n\n\\[N(A),\\qquad \\operatorname{Null}(A),\\qquad \\ker A.\\]\n\nHere \\(N(A)\\subseteq\\mathbb F^n\\), whereas the image of \\(A\\) lies in \\(\\mathbb F^m\\). (maths.tcd.ie)\n\nThe zero vector always belongs to the null space. If \\(u,v\\in\\ker T\\) and \\(a,b\\in\\mathbb F\\), linearity gives\n\n\\[T(au+bv)=aT(u)+bT(v)=0.\\]\n\nThus it is closed under linear combinations. A null space containing only the zero vector is called trivial; it is not an empty set. A linear map is an injective function precisely when its null space is trivial, because \\(T(u)=T(v)\\) is equivalent to \\(u-v\\in\\ker T\\). (maths.tcd.ie)"
    },
    {
      "heading": "Nullity and rank",
      "text": "The dimension of the null space is called the nullity. If \\(V\\) is finite-dimensional, the rank–nullity theorem states\n\n\\[\\dim V=\\dim\\ker T+\\dim\\operatorname{im}T.\\]\n\nConsequently, if \\(A\\) has \\(n\\) columns and rank \\(r\\),\n\n\\[\\operatorname{nullity}(A)=n-r.\\]\n\nNullity therefore counts the independent input directions lost under the transformation. (homepages.ucl.ac.uk)\n\nA matrix has a trivial null space exactly when its columns are linearly independent. In particular, an \\(m\\times n\\) matrix with \\(n>m\\) necessarily has a nontrivial null space: its rank cannot exceed \\(m\\). This conclusion concerns homogeneous systems; it does not establish that an arbitrary equation \\(Ax=b\\) has a solution. (live.ocw.mit.edu)"
    },
    {
      "heading": "Computing a basis",
      "text": "For exact calculations, Gaussian elimination reduces \\(A\\) to row-echelon form. Elementary row operations preserve the solutions of \\(Ax=0\\), so they preserve its null space. Pivot variables are expressed in terms of free variables. Setting one free variable to \\(1\\) and the others to \\(0\\), successively, produces a basis of the null space. The number of resulting basis vectors equals the number of nonpivot columns. (ocw.mit.edu)\n\nFor example, take\n\n\\[A=\\begin{pmatrix} 1&2&3\\\\ 2&4&6 \\end{pmatrix}.\\]\n\nThe second equation duplicates the first, leaving\n\n\\[x_1+2x_2+3x_3=0.\\]\n\nWriting \\(x_2=s\\) and \\(x_3=t\\) gives\n\n\\[x=s\\begin{pmatrix}-2\\\\1\\\\0\\end{pmatrix} +t\\begin{pmatrix}-3\\\\0\\\\1\\end{pmatrix}.\\]\n\nHence\n\n\\[N(A)=\\operatorname{span}\\{(-2,1,0)^T,(-3,0,1)^T\\}.\\]\n\nThis calculated example has rank \\(1\\) and nullity \\(2\\). Its null space is a plane through the origin in \\(\\mathbb R^3\\), although its displayed basis is not unique."
    },
    {
      "heading": "Geometry and solution sets",
      "text": "For a real matrix, the null space is the orthogonal complement of the row space:\n\n\\[N(A)=\\operatorname{im}(A^T)^\\perp,\\]\n\nwhere \\(A^T\\) denotes the transpose. Each equation in \\(Ax=0\\) says that \\(x\\) is perpendicular to a row of \\(A\\), using the standard inner product. The left null space, \\(N(A^T)\\), is correspondingly perpendicular to the column space and has dimension \\(m-r\\). For complex matrices, these statements use the conjugate transpose \\(A^*\\). (ocw.mit.edu)\n\nFor a consistent system of linear equations \\(Ax=b\\), let \\(x_p\\) be one particular solution. Then every solution has the form\n\n\\[x=x_p+z,\\qquad z\\in N(A).\\]\n\nIndeed, subtracting two solutions yields a homogeneous solution, and adding any homogeneous solution preserves \\(Ax=b\\). The solution set is therefore an affine space obtained by translating the null space. It is a linear subspace only when \\(b=0\\). A consistent system has exactly one solution if and only if \\(N(A)=\\{0\\}\\). (people.cs.uchicago.edu)"
    },
    {
      "heading": "Numerical computation and applications",
      "text": "In numerical computation, singular value decomposition can construct an orthonormal basis for the null space. In a full decomposition \\(A=U\\Sigma V^*\\), the columns of \\(V\\) corresponding to zero diagonal entries of \\(\\Sigma\\), together with any additional right-side directions, span \\(N(A)\\). Computations with floating-point arithmetic usually classify sufficiently small singular values as zero using a tolerance. This produces an effective numerical null space, whose dimension may differ from the exact algebraic nullity. (docs.scipy.org)\n\nIn ordinary least squares, null-space components can be added to a coefficient vector without changing its fitted values. The Moore–Penrose pseudoinverse selects the least-squares solution of minimum Euclidean norm, which has no component in \\(N(A)\\). In a linear inverse problem, the same structure identifies changes to the unknown that cannot be distinguished by the measurements. (ocw.mit.edu)"
    }
  ],
  "links": [
    "linear-algebra",
    "linear-map",
    "kernel-linear-map",
    "matrix",
    "linear-subspace",
    "vector-space",
    "field-mathematics",
    "image-linear-map",
    "linear-combination",
    "injective-function",
    "dimension-vector-space",
    "rank-nullity-theorem",
    "matrix-rank",
    "linear-independence",
    "gaussian-elimination",
    "basis-linear-algebra",
    "orthogonal-complement",
    "matrix-transpose",
    "inner-product",
    "system-of-linear-equations",
    "affine-space",
    "singular-value-decomposition",
    "orthonormal-basis",
    "floating-point-arithmetic",
    "ordinary-least-squares",
    "moore-penrose-pseudoinverse",
    "inverse-problem"
  ],
  "sources": [
    {
      "title": "maths.tcd.ie/~mozgovoy/data/notes_LA2.pdf",
      "url": "https://www.maths.tcd.ie/~mozgovoy/data/notes_LA2.pdf"
    },
    {
      "title": "homepages.ucl.ac.uk/~ucahmto/0007/_book/3-3-linear-maps.html",
      "url": "https://www.homepages.ucl.ac.uk/~ucahmto/0007/_book/3-3-linear-maps.html"
    },
    {
      "title": "live.ocw.mit.edu/courses/18-06-linear-algebra-spring-2010/c1729bd4d02104d1632a4…",
      "url": "https://live.ocw.mit.edu/courses/18-06-linear-algebra-spring-2010/c1729bd4d02104d1632a4dd8cc63d055_MIT18_06S10_L11.pdf"
    },
    {
      "title": "ocw.mit.edu/courses/18-06sc-linear-algebra-fall-2011/b547dd4153ccc350860df8f430…",
      "url": "https://ocw.mit.edu/courses/18-06sc-linear-algebra-fall-2011/b547dd4153ccc350860df8f43085c2d6_MIT18_06SCF11_Ses1.14sum.pdf"
    },
    {
      "title": "ocw.mit.edu/courses/18-085-computational-science-and-engineering-i-fall-2008/75…",
      "url": "https://ocw.mit.edu/courses/18-085-computational-science-and-engineering-i-fall-2008/7549c43ec4a5c4544fb68cb626396fbd_nutshell.pdf"
    },
    {
      "title": "people.cs.uchicago.edu/~laci/HANDOUTS/linalgbook.pdf",
      "url": "https://people.cs.uchicago.edu/~laci/HANDOUTS/linalgbook.pdf"
    },
    {
      "title": "docs.scipy.org/doc/scipy/reference/generated/scipy.linalg.null_space.html",
      "url": "https://docs.scipy.org/doc/scipy/reference/generated/scipy.linalg.null_space.html"
    },
    {
      "title": "ocw.mit.edu/courses/12-864-inference-from-data-and-models-spring-2005/76f046407…",
      "url": "https://ocw.mit.edu/courses/12-864-inference-from-data-and-models-spring-2005/76f0464070f454ed5f7477601bf6e8ac_least_square_2_4.pdf"
    }
  ],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}