# Missing $inserted error in latex Hi I'm trying to prepare for my presentation, but I have a problem with my latex code. I'm sure that I'm using "amsmath" package. I tried to fix this, but still have a lot of errors. \begin{frame}{ZIP regression models with covariates} \begin{itemize} \item Regression-type models can adjust for covariate effects and assess relationships between key predictors and the response \item Covariates enter ZIP regression model at both the Bernoulli zero-inflation and Poison count stages\\$\rar$2 sets of parameters corresponding to p and$\lambda$\\ ~~~~~~~~ $$\begin{cases} \(\lambda$$ : loglinear model $$\rar log(\frac{p}{1-p}) = \textbf{X_1\alpha} = \alpha_0 + \alpha_1X_{11} + \alpha_2X_{12} + \dots + \alpha_mX_{1m}$$\\ p : logit model $$\rar log(\lambda) = \textbf{X_2\beta} = \beta_0 + \beta_1X_{21} + \beta_2X_{22} + \dots + \beta_lX_{2l}$$ \end{cases}\) \vspace{0.2cm} \item $$\boldsymbol{X_1} = (1, X_{11}, X_{12}, \dots, X_{1m})$$ : covariate vector included in the zero stage\\ $$\boldsymbol{X_2} = (1, X_{21}, X_{22}, \dots, X_{2l})$$ : covariate vector included in the Poisson stage \item $$\boldsymbol{\alpha} = (\alpha_0, \alpha_1, \dots, \alpha_m)^{T}, \boldsymbol{\beta} = (\beta_0, \beta_1, \dots, \beta_l)^{T}$$\\$\rarcorresponding coefficient vectors \end{itemize} \end{frame}  Is there something wrong? • The cases environment can not be used outpude of math mode. Commented Jul 13, 2020 at 12:27 • \rar is an undefined code I guess ? – user209604 Commented Jul 13, 2020 at 13:11 ## 2 Answers  \usepackage{amsmath} \begin{document} \begin{frame}{Zip model} \begin{itemize} \item Probability mass function of Y $$*Pr(Y = y\mid\lambda, p) = \begin{cases} p + (1 - p)e^{-\lambda}, & \text{if y = 0}\\ (1 - p)\frac{e^{-\lambda}\lambda^{y}}{y!}, & \text{if y > 0} \end{cases}$$ \item \begin{aligned}E(Y) =& (1 - p)\lambda \\ *Var(Y) =& \lambda(1 - p)(1 + p\lambda)\end{aligned} \end{itemize} \end{frame} \end{document}  You can replace $$...$$by ...$ The second Text I think I found the problem  \item$\boldsymbol{X_1} = (1, X_{11}, X_{12}, \dots, X_{1m})$: covariate vector included in the zero stage\\$\boldsymbol{X_2} = (1, X_{21}, X_{22}, \dots, X_{2l})$: covariate vector included in the Poisson stage \item$\boldsymbol{\alpha} = (\alpha_0, \alpha_1, \dots, \alpha_m)^{T}, \boldsymbol{\beta} = (\beta_0, \beta_1, \dots, \beta_l)^{T}\$\\


And :

$$\begin{cases} \lambda : \text{loglinear model}\ \log(\frac{p}{1-p}) = \textbf{X_{1\alpha}} = \alpha_0 + \alpha_{1X}_{11} + \alpha_{2X}_{12} + \dots + \alpha_{mX}_{1m} \\ p : \text{logit model} \ \log(\lambda) = \textbf{X_{2\beta}} = \beta_0 + \beta_{1X}_{21} + \beta_{2X}_{22} + \dots + \beta_{lX}_{2l} \end{cases}$$


I didn't get the \item  bcs I didn't use its package

• Hi, add always a \documentclass into your orignal code :-) Commented Jul 27, 2020 at 12:00

\documentclass{beamer}

\usepackage{amsmath}
\begin{document}
\begin{frame}{ZIP model}
\begin{itemize}
\item Probability mass function of Y
$$*Pr(Y = y\mid\lambda, p) = \begin{cases} p + (1 - p)e^{-\lambda}, & \text{if y = 0}\\ (1 - p)\frac{e^{-\lambda}\lambda^{y}}{y!}, & \text{if y} > 0 \end{cases}$$
\item \begin{aligned}E(Y) =& (1 - p)\lambda \\ *Var(Y) =& \lambda(1 - p)(1 + p\lambda)\end{aligned}
\end{itemize}
\end{frame}

\end{document}

• Thanks a lot! I'm sorry but I have one more problem :( Commented Jul 13, 2020 at 12:57
• I tried to change other codes like yours, and still I get a lot of errors... Commented Jul 13, 2020 at 12:58