initial the synthetic control object

synthetic_control(outcome = final$Violent crime, # outcome unit = final$Area, # unit index in the panel data time = final$Year, # time index in the panel data i_unit = "Colorado", # unit where the intervention occurred i_time = 2013, # time period when the intervention occurred generate_placebos=TRUE # generate placebo synthetic controls (for inference) ) %>%

Generate the aggregate predictors used to fit the weights

average log income, retail price of cigarettes, and proportion of the

population between 15 and 24 years of age from 1980 - 1988

generate_predictor(time_window = 2013:2019, tx_desemprego = mean(Taxa de desemprego, na.rm = T), renda_percapita = mean(Renda per Capita, na.rm = T), tx_conservadorismo = mean(Conservadorismo, na.rm = T)) %>%

Generate the fitted weights for the synthetic control

generate_weights(optimization_window = 2007:2019, # time to use in the optimization task margin_ipop = .02,sigf_ipop = 7,bound_ipop = 6 # optimizer options ) %>%

Generate the synthetic control

generate_control()

Error in dplyr::mutate(): ℹ In argument: placebo = ifelse(final$Area != i_unit, 1, 0). Caused by error: ! placebo must be size 7 or 1, not 91. Run rlang::last_error() to see where the error occurred.

This error is happening and i've runned out of ideas on how to fix it. Can someone help me, i don't know what else to do.

Looks like the issue is probably with your inputs. tidysynth uses quasiquotation (standard tidyverse rlang stuff) so you don't need/shouldn't index on the variable as you're doing.

Try:

final %>% 
synthetic_control(
        outcome =`Violent crime`, 
        unit = Area, # unit index in the panel data
        time = Year, # time index in the panel data
        i_unit = "Colorado", # unit where the intervention occurred
        i_time = 2013, # time period when the intervention occurred
        generate_placebos=TRUE # generate placebo synthetic controls (for inference)
) %>%
    # and so on ...

Also, FWIW, you should opt for no spaces in your variable naming conventions. It'll make your life way easier. (e.g. Violent crime should be Violent_crime).

I hope that helps!

edunford / tidysynth