Bases | ️ Base16 and Base32 encoding in Swift

Scratching my head on this one. I've simply added a new field to a model.

I am working on the Kaggle: Abalone dataset and I am facing a weird problem when plotting a boxplot.

I have a requirement to build a SSIS package that sends HTML formatted emails and then saves the emails as tiff files. I have created a script task that processes the necessary records and then coverts the HTML code to the tiff. I have split the process into separate packages, the email send works fine the converting HTML to tiff is causing the issue.

When running the package manually it will process all files without any issues. my test currently is about 315 files this needs to be able to process at least 1,000 when finished with the ability to send up to 10,000 at one time. The problem is when I set the package to execute using SQL Server Agent it stops at 207 files. The package is deployed to SQL Server 2019 in the SSIS Catalog

What I have tried so far

I started with the script being placed in a SSIS package and deployed to the server and calling the package from a step (works 99.999999% of the time with all packages) tried both 32 and 64 bit runtime. Never any error messages just Unexpected Termination when looking at the execution reports. When clicking in the catalog and executing package it will process all the files. The SQL Server Agent is using a proxy and I also created another proxy account with my admin credentials to test for any issues with the account.

Created another package to call the package and used the Execute Package Task to call the first package, same result 207 files. Changed the execute Process task to an Execute SQL Task and tried the script that is created to manually start a package in the catalog 207 files. Tried executing the script from the command line both through the other SSIS package and the SQL Server Agent directly same results 207 files. If I try any of those methods directly outside SQL Server Agent the process runs no issues.

I converted the script task to a console application and it works processing all the files. When calling the executable file from any method from the SQL Server Agent it once again stops at the 207 files.

I have consulted with the companies DBA and Systems teams and they have not found anything that could be causing this error. There seems to be some type of limit that no matter the method of execution SQL Server Agent will not allow. I have mentioned looking at third-party applications but have been told no.

I have included the code below that I have been able to piece together. I am a SQL developer so C# is outside my knowledge base. Is there a way to optimize the code so it only uses one thread or does a cleanup between each letter. There may be a need for this to create over ten thousand letters at certain times.

Update

I have replaced the code with the new updated code. The email and image creation are all included as this is what the final product must do. When sending the emails there is a primary and secondary email address and depending on what email address is used it will change what the body of the email contains. When looking at the code there is a section of try catch that sends to primary when indicated to and if that fails it send to secondary instead. I am guessing there is a much cleaner way of doing that section but this is my first program as I work in SQL for everything else.

Thank You for all the suggestions and help.

Updated Code

I am having really weird issues with memory leak and I cannot find source after long hours of testing, debugging and trial/error attempts.

The source of memory leak is reloading object base on push from subject. On which part I load current database values and add it to SourceCache.

This part is getting the push, to initiate reload

I am trying to link Cortex-M4 firmware with clang + lld. The gcc build works fine. I am using the stock CMSIS linker script with only RAM & ROM size adjusted (bases are the same). Beginning of the script (without comments):

I am new to angular and working on reactive form.

I have a html table in which by looping through I have generated controls

I want to add validation based on following cases

1. When this page loads than by default Save button should be disabled( which I have achieved by used [disabled]="!myform.valid"
2. Save button should enable only when user enter value in any of text boxes or select check box in that particular row. Checkbox select and add value in text box should not allow to user. Either checkbox should be selected or user can enter value in any of text-boxes.

I tried this to achieve

I am using MongoDB(Mongo Atlas) in my Django app. All was working fine till yesterday. But today, when I ran the server, it is showing me the following error on console

I'm trying to implement a gradient-free optimizer function to train convolutional neural networks with Julia using Flux.jl. The reference paper is this: https://arxiv.org/abs/2005.05955. This paper proposes RSO, a gradient-free optimization algorithm updates single weight at a time on a sampling bases. The pseudocode of this algorithm is depicted in the picture below.

optimizer_pseudocode

I'm using MNIST dataset.

Unambiguous DNA sequences consist only of the nucleobases adenine (A), cytosine (C), guanine (G), thymine (T). For human consumption, the bases may be represented by the corresponding char in either uppercase or lowercase: A, C, G, T, or a, c, g, t. This representation is inefficient, however, when long sequences need to be stored. Since only four symbols need to be stored, each symbol can be assigned a 2-bit code. The commonly used .2bit-format specified by UCSC does exactly that, using the following encoding: T = 0b00, C = 0b01, A = 0b10, G = 0b11.

The C code below shows a reference implementation written for clarity. Various open-source software that converts genomic sequences represented as a char sequence typically uses a 256-entry lookup table indexed by each char in sequence. This also isolates from the internal representation of char. However, memory access is energetically expensive, even if the access is to an on-chip cache, and generic table look-ups are difficult to SIMDize. It would therefore be advantageous if the conversion could be accomplished by simple integer arithmetic. Given that ASCII is the dominating char encoding, one can restrict to that.

What are efficient computational approaches to convert nucleobases given as ASCII characters to their .2bit-representation?