The missMDA package in R is designed to perform principal component analysis (PCA), multiple correspondence analysis (MCA), factorial analysis for mixed data (FAMD), and multiple factor analysis (MFA) on datasets with missing values. It allows for the estimation of parameters such as scores, loadings, and graphical representations despite missing data. The package also supports single and multiple imputation methods for handling incomplete data involving continuous, categorical, and mixed variables. In the PCA framework, confidence areas around graphical outputs represent variability across different imputations, helping assess the credibility of results from incomplete datasets. The package addresses the issue of missing values by integrating imputation with parameter estimation. It provides a broad range of applications for imputing incomplete datasets and applying statistical methods. The method is based on singular value decomposition (SVD) and is suitable for both continuous and categorical data. The package includes a multiple imputation method that allows for the calculation of confidence areas on graphical outputs to assess the credibility of results from incomplete data. The missMDA package also handles missing values in MCA by extending PCA methods to categorical data. It uses an iterative algorithm to impute missing values in the indicator matrix and updates margins accordingly. A regularized version of the algorithm is also available to mitigate overfitting issues. For FAMD, the package performs PCA on a weighted matrix that balances the influence of continuous and categorical variables. The method involves standardizing continuous variables and normalizing dummy variables based on the proportion of individuals in each category. The algorithm iteratively imputes missing values and updates means, standard deviations, and margins until convergence. The missMDA package extends PCA methods to multi-table data through MFA. It handles datasets with multiple tables by performing PCA on each table and then combining the results. The package provides functions for estimating the number of dimensions, imputing missing values, and performing analyses on imputed data. The results are visualized using graphical outputs that show the positions of individuals and variables, with confidence areas indicating variability across different imputations. The package is implemented in R and is available for use in various fields, including agriculture, biology, and social sciences. It provides a comprehensive framework for handling missing values in multivariate data analysis and supports both single and multiple imputation methods. The package is designed under the assumption that missing values are missing at random (MAR), and it does not address the case of missing non-at random (MNAR) values. The methods are validated using real datasets and are suitable for a wide range of applications involving incomplete data.The missMDA package in R is designed to perform principal component analysis (PCA), multiple correspondence analysis (MCA), factorial analysis for mixed data (FAMD), and multiple factor analysis (MFA) on datasets with missing values. It allows for the estimation of parameters such as scores, loadings, and graphical representations despite missing data. The package also supports single and multiple imputation methods for handling incomplete data involving continuous, categorical, and mixed variables. In the PCA framework, confidence areas around graphical outputs represent variability across different imputations, helping assess the credibility of results from incomplete datasets. The package addresses the issue of missing values by integrating imputation with parameter estimation. It provides a broad range of applications for imputing incomplete datasets and applying statistical methods. The method is based on singular value decomposition (SVD) and is suitable for both continuous and categorical data. The package includes a multiple imputation method that allows for the calculation of confidence areas on graphical outputs to assess the credibility of results from incomplete data. The missMDA package also handles missing values in MCA by extending PCA methods to categorical data. It uses an iterative algorithm to impute missing values in the indicator matrix and updates margins accordingly. A regularized version of the algorithm is also available to mitigate overfitting issues. For FAMD, the package performs PCA on a weighted matrix that balances the influence of continuous and categorical variables. The method involves standardizing continuous variables and normalizing dummy variables based on the proportion of individuals in each category. The algorithm iteratively imputes missing values and updates means, standard deviations, and margins until convergence. The missMDA package extends PCA methods to multi-table data through MFA. It handles datasets with multiple tables by performing PCA on each table and then combining the results. The package provides functions for estimating the number of dimensions, imputing missing values, and performing analyses on imputed data. The results are visualized using graphical outputs that show the positions of individuals and variables, with confidence areas indicating variability across different imputations. The package is implemented in R and is available for use in various fields, including agriculture, biology, and social sciences. It provides a comprehensive framework for handling missing values in multivariate data analysis and supports both single and multiple imputation methods. The package is designed under the assumption that missing values are missing at random (MAR), and it does not address the case of missing non-at random (MNAR) values. The methods are validated using real datasets and are suitable for a wide range of applications involving incomplete data.