Monday, February 13, 2012
My long suffering editor (who puts up with me for some unknown reason, lol) reviewed the pattern for clarity. It turns out that most editors do this as well, and leave the issue of numbers to the test knitters. I thought she would go over them to double check before we sent the pattern to the testers. So, that was my error. Thankfully, my testers (who are all at different points in the pattern) have found some of the bigger problems and they were easy to address. But I discovered that my sleeve numbers were off in the 34" size, and my fronts were off in my 38" sample, so I realized that I needed to go rework the numbers for a 5th time. That photo you see above...that represents my latest round of number crunching.
In a particularly down moment, where I doubted my ability to pull this design together and seriously considered scrapping it (and I probably would have, if it weren't for the fact that I'm doing this in public and pride won't let me, lol), I thought "what if I just reverse the whole thing and design it from the bottom-up?" I've done that before successfully. But some of you out there expressed that you really, really wanted it to remain top down. So what's a girl to do? Work the numbers from the bottom up! Start with the final numbers and work backward like a word problem! Why the heck didn't I think of that sooner? (Probably because I didn't pray hard enough, lol).
I started with the finished numbers for the body and the sleeves. I pulled out my Handy Book of Sweaters again for help in determining the standard depth for a raglan sweater. Then, I took the body numbers and made sure they were divisible by 4 (to accomodate my ribbing), and split that number in half to give me a front and a back. I took the front and split that in half again, so I would know how many sts I will end up with on either side of the v-neck. From there, I used my formula from last time to determine how many stitches would be decreased when I worked the neck and sleeve shaping. I subtracted that number from the total sts where the v-neck joins, and I got the number of sts to cast on. Genius, and much faster than pulling numbers out of thin air and trying them out till I found the right one. Knowing the raglan depth also provided me with the number of total rows through the yoke, so I knew how many decrease rows I would have, and how many stitches to decrease for the sleeves and the back. A little subtraction, and I had those cast on numbers too.
The next step was to write all the cast on numbers down, and add them together for the total number of stitches to cast on. Now, here's where it got really complicated last time. I needed to move some of the sts around so that I could fit the cable into the raglan. Shirley Paden's book, Knitware Design Workshop mentions that the cross-back should be about one inch shorter than the front, so that the garment sits correctly. Using that logic, I subtracted 4 sts from the back-neck numbers and moved them to the fronts. This accomplished my goal of providing enough sts to fit the raglan in, and creating a cross-back that is one inch shorter than the front. To make sure I did everything correctly, I went down my paper listing the bust sizes, and went across the paper listing the different sections of the cast on (front, sleeve, back, sleeve, front) and the total. I plotted the numbers for every size. Then I went back and moved the stitches around, plugging the cable into the fronts and the back (front, raglan, sleeve, raglan, back, raglan, sleeve, raglan, front). I added those numbers again to make sure the total remained unchanged. Finally, I listed the total number of the front sts and back sts for each size, along the bottom of the page. I verified that the fronts had 4 more sts than the backs, and that the total of each size matched the total body numbers that I started with. It all worked out, and my editor said that they all look good to her as well.
So what does that mean for me? Well, it means that I need to completely adjust the cast-on numbers in my pattern, and that my testers will all have different sleeve numbers that I originally thought. That won't really be an issue, because I can help them adjust the sleeves to fit correctly so their finished piece will be wearable. But it definitely means that I'll need a second round of testing. It also means that I had to rip back my sample and start one more time with the latest set of numbers. And lastly, it means that I'm definitely taking Rock + Purl's grading class this August, because there's no way I'm going to deal with this mess again without a little more training :)