Linear Response for Contracting on Average Iterated Function Systems
Jianning Fu · 2026
Read Linear Response for Contracting on Average Iterated Function Systems by Jianning Fu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-13111.txt.
Introduction
Fu studies linear response for a random contracting-on-average iterated function system with two maps, contraction ratios spanning one, and product below one. He asks whether the integral of a test function against the steady measure varies smoothly with the first contraction ratio. He proves three cases where it is differentiable, with the derivative matching the formal derivative, and gives enough conditions for a smooth bounded test function making it non-differentiable.
Catalog record
| CREATOR | Jianning Fu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 131343 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.13111
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Linear Response for Contracting on Average Iterated Function Systems Jianning Fu Abstract . Consider the following probabilistic contracting on average iterated function system Φ = { f i ( x ) = λ i x + d i , i = 1 , 2 ; p = ( 1 2 , 1 2 ) } , \Phi=\left\{f_{i}(x)=\lambda_{i}x+d_{i},\;i=1,2;\;\;p=\left(\frac{1}{2},\frac{1}{2}\right)\right\}, where the contraction ratios λ 1 , λ 2 \lambda_{1},\lambda_{2} are such that 0 < λ 1 < 1 < λ 2 0<\lambda_{1}<1<\lambda_{2} and λ 1 λ 2 < 1 \lambda_{1}\lambda_{2}<1 . Denote by μ λ 1 , λ 2 \mu_{\lambda_{1},\lambda_{2}} its stationary measure. We study the differentiability of ( ♡ ) λ 1 ↦ ∫ ℝ ϕ ( x ) d μ λ 1 , λ 2 ( x ) , (\heartsuit)\quad\quad\quad\quad\quad\lambda_{1}\mapsto\int_{\mathbb{R}}\phi(x)\,d\mu_{\lambda_{1},\lambda_{2}}(x), where ϕ \phi is a suitable test function. We establish three cases where ( ♡ ) (\heartsuit) is differentiable and show the derivative coincides with the one obtained by taking formal derivative, which can be generalized to the case of multiple maps with different probabilities. We also present sufficient conditions under which there exists a smooth, bounded test function ϕ \phi so that ( ♡ ) (\heartsuit) is not differentiable. 1. Introduction and Statement of Results To state the main results, let us first recall some basic facts of self-similar sets and measures. Let m\geq 2 and d\geq 1 be positive integers and consider m maps \mathcal{F}=(f_{1},\cdots f_{m}) acting on \mathbb{R}^{d} , with each map f_{i} having a contraction ratio 0<r_{i}<1 . That is, \forall x,y\in\mathbb{R}^{d},\quad||f_{i}(x)-f_{i}(y)||=r_{i}||x-y||,\quad i=1,\cdots m. Such a tuple of maps \mathcal{F} is called a self-similar iterated function system (self-similar IFS or IFS for short). It is well known [ 12 ] that there exists a unique compact set K=K_{\mathcal{F}} , called the attractor of \mathcal{S} , such that K=\bigcup_{i=1}^{m}f_{i}(K). Throughout, we always assume that not all f_{i} ’s share the same fixed point, so that the attractor of \mathcal{S} is not a single point. One can view the attractor via symbolic dynamics as follows. Define \mathscr{A}=\{1,\cdots m\} to be the alphabet, \Sigma=\mathscr{A}^{\mathbb{N}} the shift space and \sigma:\Sigma\to\Sigma the left shift. Consider the map \Pi=\Pi_{\mathcal{S}}:\Sigma\to\mathbb{R}^{d} that sends any word i=(i_{1},i_{2},\cdots)\in\Sigma to \Pi(i)=\lim_{n\to\infty}f_{i_{1}}\circ\cdots f_{i_{n}}(0)=:\lim_{n\to\infty}f_{i_{1}\cdots i_{n}}(0). Note that since each f_{i} is contracting, \displaystyle\lim_{n\to\infty}f_{i_{1}}\circ\cdots f_{i_{n}}(x) exists and is independent of the choice of x , justifying the above definition. The attractor K is exactly the image of the projection \Pi . An additional probability structure can be put on IFS. Namely, if \mathcal{F}=(f_{1},\cdots,f_{m}) is a self-similar IFS and p=(p_{1},\cdots,p_{m}) is a probability vector, then we call \Phi=(\mathcal{F},p) a probabilistic IFS . A Borel probability measure \mu on \mathbb{R^{d}} is called the stationary measure for the IFS \Phi , if \mu=\sum_{j=1}^{m}p_{j}(f_{j})_{*}\mu. That is, if we consider the Markov process where for any x\in\mathbb{R}^{d} , it is mapped to f_{j}(x) with probability p_{j} , then \mu is the corresponding stationary measure. It is well known that such measure is unique, so it is indeed “the” stationary measure. Moreover, it can also be described symbolically. Define \nu=p^{\mathbb{N}} to be the Bernoulli measure on \Sigma . Then \displaystyle\mu=\Pi_{*}\nu. We refer to [ 5 ] , [ 8 ] for the aforementioned properties as well as more discussions on (probabilistic) IFS. The notion of attractor and stationary measure can be generalized to the case where not all r_{i} are less than 1 . However, we require that the IFS contracts on average , meaning that \sum_{j=1}^{m}p_{j}\log r_{j}<0. In this case, there is also a unique invariant measure \mu [ 13 ] , which is still characterized by \mu=\Pi_{*}\nu . One slight difference is that in the contracting on average case, the projection \Pi is only defined for \nu -almost every words. In addition, its support is no longer compact. Many results in the contracting case can be generalized to this stationary measure, including but not restricted to its dimension, density, order of tail and absolute continuity (with respect to Hausdorff measure of suitable dimension) [ 11 ] [ 16 ] [ 9 ] [ 14 ] [ 15 ] [ 17 ] . In the present work, we study the linear response problem for stationary measures of a contracting on average IFS. In order to keep the notation simple, we restrict ourselves to the following system \Phi=\left\{f_{i}(x)=\lambda_{i}x+d_{i},\;i=1,2;\;\;p=\left(\frac{1}{2},\frac{1}{2}\right)\right\}, where the contraction ratios \lambda_{1},\lambda_{2} are such that 0<\lambda_{1}<1<\lambda_{2} and \lambda_{1}\lambda_{2}<1 . We discuss in Section 7 more general setting where our approach works. Denote by \mu_{\lambda_{1},\lambda_{2}} the stationary measure of this IFS. Consider the map (1.1) h=h_{\phi}:\epsilon\mapsto\int\phi\,d\mu_{\lambda_{1}+\epsilon,\lambda_{2}}, defined for \epsilon satisfying (\lambda_{1}+\epsilon)\lambda_{2}<1 , where \phi is a function of a certain regularity. By linear response theory, we mean studying sufficient regularity conditions on \phi so that h_{\phi}^{\prime}(0) exists and calculating the value of the derivative. For many smooth or piecewise smooth dynamical systems, there is a well-developed linear response theory, originated in Ruelle’s work [ 18 ] and developed by many others including but not restricted to [ 7 , 10 , 2 , 3 ] . There are also examples of failure of linear response, see for example [ 4 , 1 , 6 ] . However, not much is known about the analogous question for contracting (on average) IFS, and it is tempting to develop an analogous theory. Our method is based on the tail property of \mu_{\lambda_{1},\lambda_{2}} and large deviation theory, and may work in a more general setting. In contrast, in a separate work we will show that for Bernoulli convolutions, to get (almost everywhere) differentiability of h_{\phi} we only need Hölder continuous observable \phi . Moreover, differentiability with Hölder observable is related to the dimension of the measure. To state the main results we need additional notation. By the symbolic characterization, one get the symbolic space \Sigma=\{1,2\}^{\mathbb{N}} equipped with the probability measure \nu=(\frac{1}{2},\frac{1}{2})^{\mathbb{N}} and the projection \Pi(i_{1},i_{2},\cdots)=\sum_{j=0}^{\infty}d_{i_{j}}\lambda_{i_{1}}\cdots\lambda_{i_{j-1}}=:X. Moreover, \mu_{\lambda_{1},\lambda_{2}}=\Pi_{*}\nu will be the law of the random variable X . If \epsilon is sufficiently small so that (\lambda_{1}+\epsilon)\lambda_{2}<1 ,the random variable X(\epsilon):=\sum_{j=0}^{\infty}d_{i_{j}}\lambda_{i_{1}}(\epsilon)\cdots\lambda_{i_{j-1}}(\epsilon) is well-defined almost everywhere, where \lambda_{1}(\epsilon)=\lambda_{1}+\epsilon and \lambda_{2}(\epsilon)=\lambda_{2} , and has law \mu_{\lambda_{1}+\epsilon,\lambda_{2}} . With these notations, we can write \int_{\mathbb{R}}\phi\,d\mu_{\lambda_{1}+\epsilon,\lambda_{2}}=\int_{\Sigma}\phi(X(\epsilon))\,d\nu. Since X(\epsilon) is a formal power series in \epsilon , one can take derivative formally. It is therefore natural to ask whether (1.2) h^{(l)}(0):=\frac{d^{l}h}{d\epsilon^{l}}(0)=\int_{\Sigma}\frac{d^{l}}{d\epsilon^{l}}(\phi\circ X(\epsilon))|_{\epsilon=0}\,d\nu,\;l=1,2,\cdots under suitable assumptions on the test function \phi and \lambda_{1},\lambda_{2} . We will show three instances of equality ( 1.2 ) being true,. On the other hand, if \mu_{\lambda_{1},\lambda_{2}} is too concentrated on some intervals with small Lebesgue measure, it is possible to construct a C^{\infty} , bounded test function \phi whose derivative is large around these points, so that h_{\phi} is not differentiable at 0 and, in particular, equation ( 1.2 ) fails. We will give two sufficient conditions on \lambda_{1},\lambda_{2} for such \phi to exist. Now we state the main results. We start with three cases where equation ( 1.2 ) is true. Theorem 1.1 . Suppose r\in\mathbb{N} and 0<\lambda_{1}<1<\lambda_{2} are such that \displaystyle\frac{\lambda_{1}^{r}+\lambda_{2}^{r}}{2}<1 . Then for any function \phi\in C^{r}(\mathbb{R}) with ||\phi^{(r)}||_{C^{0}}<\infty , the average h(\epsilon):=\int\phi\,d\mu_{\lambda_{1}+\epsilon,\lambda_{2}} is differentiable at 0 up to order r and the derivatives are given by \displaystyle h^{(l)}(0) \displaystyle=\int_{\Sigma}\frac{\,d^{l}}{\,d\epsilon^{l}}(\phi\circ X(\epsilon))|_{\epsilon=0}\,d\nu (1.3) \displaystyle=\int_{\Sigma}\sum_{k_{1}+2k_{2}+\cdots lk_{l}=l}L(k_{1},\cdots,k_{l})\phi^{(k)}(X)\prod_{j=1}^{l}(X^{(j)})^{k_{j}}\,d\nu,\;l=1,2,\cdots,r. The second equality above is simply the Faa