<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix"><br>
      Ha.  Yes we certainly don't hold back from making the messages as
      long and as helpful as possible.  If the code knows, or can know
      what exactly is wrong, it's a deliberate policy to put that info
      right there into the message. data.table is written by users; i.e.
      we wrote it for ourselves doing real jobs. I think that may be the
      root of that.  If any messages could more helpful,  those
      suggestions are very welcome.<br>
      <br>
      Matt<br>
      <br>
      On 12/02/14 17:58, John Laing wrote:<br>
    </div>
    <blockquote
cite="mid:CAA3Wa=u2PwRW_Od4XTodtZ4uDSN7RZmOs33Uaaix8GN7OuRpag@mail.gmail.com"
      type="cite">
      <div dir="ltr">Thanks, Matt! With a slight amendment that works
        great:
        <div>
          <div>for (x in c("foo", "bar", "qux")) set(fbq, which(<a
              moz-do-not-send="true" href="http://is.na">is.na</a>(fbq[[x]])),

            x, FALSE)</div>
        </div>
        <div><br>
        </div>
        <div>Which highlights an opportunity to say that I really
          appreciate the unusually helpful error messages in this
          package.</div>
        <div><br>
        </div>
        <div>-John</div>
      </div>
      <div class="gmail_extra"><br>
        <br>
        <div class="gmail_quote"> On Wed, Feb 12, 2014 at 12:44 PM, Matt
          Dowle <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:mdowle@mdowle.plus.com" target="_blank">mdowle@mdowle.plus.com</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div bgcolor="#FFFFFF" text="#000000">
              <div><br>
                Hi John,<br>
                <br>
                In examples like this I'd use set() and [[,  since it's
                a bit easier to write but memory efficient too.<br>
                <br>
                for (x in c("foo", "bar", "qux"))   set(fbq, <a
                  moz-do-not-send="true" href="http://is.na"
                  target="_blank">is.na</a>(fbq[[x]]), x,
                FALSE)           [untested]<br>
                <br>
                A downside here is one repetition of the "fbq" symbol, 
                but can live with that.  If you have a large number of
                columns  (and I've been surprised just how many columns
                some poeple have!) then calling set() many times has
                lower overhead than DT[, :=],  see ?set.   Note also
                that [[ is base R, doesn't copy the column and often
                useful to use with data.table.<br>
                <br>
                Or, use get() in either i or j rather than eval().<br>
                <br>
                HTH, Matt
                <div>
                  <div class="h5"><br>
                    <br>
                    <br>
                    On 12/02/14 17:24, John Laing wrote:<br>
                  </div>
                </div>
              </div>
              <blockquote type="cite">
                <div>
                  <div class="h5">
                    <div dir="ltr">Let's say I merge together several
                      data.tables such that I wind up<br>
                      with lots of NAs:<br>
                      <br>
                      require(data.table)<br>
                      foo <- data.table(k=1:4, foo=TRUE, key="k")<br>
                      bar <- data.table(k=3:6, bar=TRUE, key="k")<br>
                      qux <- data.table(k=5:8, qux=TRUE, key="k")<br>
                      fbq <- merge(merge(foo, bar, all=TRUE), qux,
                      all=TRUE)<br>
                      print(fbq)<br>
                      #    k  foo  bar  qux<br>
                      # 1: 1 TRUE   NA   NA<br>
                      # 2: 2 TRUE   NA   NA<br>
                      # 3: 3 TRUE TRUE   NA<br>
                      # 4: 4 TRUE TRUE   NA<br>
                      # 5: 5   NA TRUE TRUE<br>
                      # 6: 6   NA TRUE TRUE<br>
                      # 7: 7   NA   NA TRUE<br>
                      # 8: 8   NA   NA TRUE<br>
                      <br>
                      I want to go through those columns and turn each
                      NA into FALSE. I can<br>
                      do this by writing code for each column:<br>
                      <br>
                      fbq.cp <- copy(fbq)<br>
                      fbq.cp[<a moz-do-not-send="true"
                        href="http://is.na" target="_blank">is.na</a>(foo),


                      foo:=FALSE]<br>
                      fbq.cp[<a moz-do-not-send="true"
                        href="http://is.na" target="_blank">is.na</a>(bar),


                      bar:=FALSE]<br>
                      fbq.cp[<a moz-do-not-send="true"
                        href="http://is.na" target="_blank">is.na</a>(qux),


                      qux:=FALSE]<br>
                      print(fbq.cp)<br>
                      #    k   foo   bar   qux<br>
                      # 1: 1  TRUE FALSE FALSE<br>
                      # 2: 2  TRUE FALSE FALSE<br>
                      # 3: 3  TRUE  TRUE FALSE<br>
                      # 4: 4  TRUE  TRUE FALSE<br>
                      # 5: 5 FALSE  TRUE  TRUE<br>
                      # 6: 6 FALSE  TRUE  TRUE<br>
                      # 7: 7 FALSE FALSE  TRUE<br>
                      # 8: 8 FALSE FALSE  TRUE<br>
                      <br>
                      But I can't figure out how to do it in a loop.
                      More precisely, I can't<br>
                      figure out how to make the [ operator evaluate its
                      first argument in<br>
                      the context of the data.table. All of these have
                      no effect:<br>
                      for (x in c("foo", "bar", "qux")) fbq[<a
                        moz-do-not-send="true" href="http://is.na"
                        target="_blank">is.na</a>(x), eval(x):=FALSE]<br>
                      for (x in c("foo", "bar", "qux")) fbq[<a
                        moz-do-not-send="true" href="http://is.na"
                        target="_blank">is.na</a>(eval(x)),
                      eval(x):=FALSE]<br>
                      for (x in c("foo", "bar", "qux")) fbq[eval(<a
                        moz-do-not-send="true" href="http://is.na"
                        target="_blank">is.na</a>(x)), eval(x):=FALSE]<br>
                      <br>
                      I'm running R 3.0.2 on Linux, data.table 1.8.10.<br>
                      <br>
                      Thanks in advance,<br>
                      John</div>
                    <br>
                    <fieldset></fieldset>
                    <br>
                  </div>
                </div>
                <pre>_______________________________________________
datatable-help mailing list
<a moz-do-not-send="true" href="mailto:datatable-help@lists.r-forge.r-project.org" target="_blank">datatable-help@lists.r-forge.r-project.org</a>
<a moz-do-not-send="true" href="https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help" target="_blank">https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help</a></pre>
              </blockquote>
              <br>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>